From 9deab9e9adfacbb657643f2d0c4754a2831a5dbe Mon Sep 17 00:00:00 2001 From: Klein <165743429+kleinpanic@users.noreply.github.com> Date: Fri, 13 Dec 2024 23:01:28 -0500 Subject: [PATCH] griffins updated files (#8) Co-authored-by: gwitthub --- src/griffin-stuff/API/API_1.ipynb | 184 -- src/griffin-stuff/API/API_2 (1).ipynb | 2074 ----------------- .../Trading_Bot_Development_Strategy (1).docx | Bin 41835 -> 0 bytes 3 files changed, 2258 deletions(-) delete mode 100644 src/griffin-stuff/API/API_1.ipynb delete mode 100644 src/griffin-stuff/API/API_2 (1).ipynb delete mode 100644 src/griffin-stuff/API/Trading_Bot_Development_Strategy (1).docx diff --git a/src/griffin-stuff/API/API_1.ipynb b/src/griffin-stuff/API/API_1.ipynb deleted file mode 100644 index 4a515ba..0000000 --- a/src/griffin-stuff/API/API_1.ipynb +++ /dev/null @@ -1,184 +0,0 @@ -{ - "cells": [ - { - "cell_type": "code", - "execution_count": 13, - "id": "69d88f26-f288-4a23-8be5-3e8317e23731", - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "ERROR -1 2104 Market data farm connection is OK:usfarm.nj\n", - "ERROR -1 2104 Market data farm connection is OK:usfuture\n", - "ERROR -1 2104 Market data farm connection is OK:cashfarm\n", - "ERROR -1 2104 Market data farm connection is OK:usfarm\n", - "ERROR -1 2106 HMDS data farm connection is OK:ushmds\n", - "ERROR -1 2158 Sec-def data farm connection is OK:secdefnj\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Historical Data Ended\n", - " Date Open High Low Close Volume\n", - "0 20241030 18:00:00 69.10 69.10 68.96 69.02 378\n", - "1 20241030 18:05:00 69.02 69.07 69.01 69.05 99\n", - "2 20241030 18:10:00 69.06 69.07 69.01 69.01 103\n", - "3 20241030 18:15:00 69.01 69.02 69.00 69.00 54\n", - "4 20241030 18:20:00 69.01 69.01 68.99 69.00 25\n", - "5 20241030 18:25:00 69.00 69.05 69.00 69.04 40\n", - "6 20241030 18:30:00 69.05 69.05 69.03 69.03 63\n", - "7 20241030 18:35:00 69.03 69.03 69.00 69.00 64\n", - "8 20241030 18:40:00 68.99 69.01 68.98 68.99 60\n", - "9 20241030 18:45:00 68.99 68.99 68.95 68.97 66\n", - "10 20241030 18:50:00 68.97 69.00 68.96 68.99 44\n", - "11 20241030 18:55:00 68.98 68.98 68.97 68.98 23\n", - "12 20241030 19:00:00 68.98 69.02 68.98 69.01 48\n", - "13 20241030 19:05:00 69.02 69.03 69.00 69.01 31\n", - "14 20241030 19:10:00 69.02 69.02 69.00 69.00 22\n", - "15 20241030 19:15:00 69.00 69.00 68.99 68.99 11\n", - "16 20241030 19:20:00 68.99 68.99 68.95 68.95 40\n", - "17 20241030 19:25:00 68.95 68.95 68.94 68.94 55\n", - "18 20241030 19:30:00 68.94 68.96 68.93 68.95 54\n", - "19 20241030 19:35:00 68.95 68.97 68.95 68.96 29\n", - "20 20241030 19:40:00 68.96 68.98 68.96 68.98 47\n", - "21 20241030 19:45:00 68.98 68.99 68.95 68.95 65\n", - "22 20241030 19:50:00 68.96 68.98 68.96 68.97 16\n", - "23 20241030 19:55:00 68.97 68.97 68.94 68.94 35\n", - "24 20241030 20:00:00 68.95 68.99 68.91 68.92 369\n", - "25 20241030 20:05:00 68.91 68.94 68.91 68.93 74\n", - "26 20241030 20:10:00 68.93 68.95 68.89 68.94 187\n", - "27 20241030 20:15:00 68.94 68.95 68.92 68.94 81\n", - "28 20241030 20:20:00 68.95 68.97 68.94 68.96 89\n", - "29 20241030 20:25:00 68.96 68.96 68.92 68.94 96\n", - "30 20241030 20:30:00 68.94 68.98 68.93 68.96 94\n", - "31 20241030 20:35:00 68.97 68.97 68.93 68.94 66\n", - "32 20241030 20:40:00 68.95 68.95 68.93 68.94 44\n", - "33 20241030 20:45:00 68.93 68.96 68.93 68.94 98\n", - "34 20241030 20:50:00 68.94 68.94 68.92 68.92 95\n" - ] - } - ], - "source": [ - "from ibapi.client import EClient\n", - "from ibapi.wrapper import EWrapper\n", - "from ibapi.contract import Contract\n", - "import threading\n", - "import time\n", - "import pandas as pd\n", - "\n", - "# Define the IB API app\n", - "class IBApi(EWrapper, EClient):\n", - " def __init__(self):\n", - " EClient.__init__(self, self)\n", - " self.data = [] # Initialize an empty list to store data\n", - "\n", - " # Override the historicalData function to process and store incoming data\n", - " def historicalData(self, reqId, bar):\n", - " # Append the data as a dictionary to self.data\n", - " self.data.append({\n", - " \"Date\": bar.date,\n", - " \"Open\": bar.open,\n", - " \"High\": bar.high,\n", - " \"Low\": bar.low,\n", - " \"Close\": bar.close,\n", - " \"Volume\": bar.volume\n", - " })\n", - "\n", - " def historicalDataEnd(self, reqId, start, end):\n", - " print(\"Historical Data Ended\")\n", - " # Convert the data to a DataFrame when data collection is complete\n", - " self.df = pd.DataFrame(self.data)\n", - " print(self.df) # Display the DataFrame to verify\n", - " self.disconnect() # Disconnect after data collection is complete\n", - "\n", - "# Define the app handler for running in the notebook\n", - "class IBApp:\n", - " def __init__(self):\n", - " self.app = IBApi()\n", - "\n", - " def connect(self):\n", - " self.app.connect(\"127.0.0.1\", 7496, 0) # Change port if needed\n", - " thread = threading.Thread(target=self.run_app, daemon=True)\n", - " thread.start()\n", - " time.sleep(1) # Allow time for the connection to establish\n", - "\n", - " def run_app(self):\n", - " self.app.run()\n", - "\n", - " def request_oil_data(self):\n", - " # Define the contract for Crude Oil Futures\n", - " contract = Contract()\n", - " contract.symbol = \"CL\"\n", - " contract.secType = \"FUT\"\n", - " contract.exchange = \"NYMEX\"\n", - " contract.currency = \"USD\"\n", - " contract.lastTradeDateOrContractMonth = \"202412\" # Example: Dec 2024 contract\n", - "\n", - " # Request historical data\n", - " self.app.reqHistoricalData(\n", - " reqId=1,\n", - " contract=contract,\n", - " endDateTime='',\n", - " durationStr='1 D', # 1 month\n", - " barSizeSetting='5 mins',\n", - " whatToShow='TRADES',\n", - " useRTH=0,\n", - " formatDate=1,\n", - " keepUpToDate=False,\n", - " chartOptions=[]\n", - " )\n", - "\n", - " def disconnect(self):\n", - " self.app.disconnect()\n", - "\n", - "# Create an instance and connect\n", - "app = IBApp()\n", - "app.connect()\n", - "\n", - "# Request data and output to a DataFrame\n", - "app.request_oil_data()\n", - "\n", - "# Wait for data retrieval to complete\n", - "time.sleep(10)\n", - "\n", - "# Access the DataFrame\n", - "df = app.app.df if hasattr(app.app, 'df') else pd.DataFrame()" - ] - }, - { - "cell_type": "code", - "execution_count": 17, - "id": "2088c621-81d3-46f0-8596-ce05d1a89fd4", - "metadata": {}, - "outputs": [], - "source": [ - "data = df.to_csv()" - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3 (ipykernel)", - "language": "python", - "name": "python3" - }, - "language_info": { - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "file_extension": ".py", - "mimetype": "text/x-python", - "name": "python", - "nbconvert_exporter": "python", - "pygments_lexer": "ipython3", - "version": "3.12.3" - } - }, - "nbformat": 4, - "nbformat_minor": 5 -} diff --git a/src/griffin-stuff/API/API_2 (1).ipynb b/src/griffin-stuff/API/API_2 (1).ipynb deleted file mode 100644 index ad37b44..0000000 --- a/src/griffin-stuff/API/API_2 (1).ipynb +++ /dev/null @@ -1,2074 +0,0 @@ -{ - "cells": [ - { - "cell_type": "code", - "execution_count": 11, - "id": "a5a771ce-8155-453a-aa48-6cc64e64f9fa", - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "ERROR -1 2104 Market data farm connection is OK:usfarm.nj\n", - "ERROR -1 2104 Market data farm connection is OK:usfuture\n", - "ERROR -1 2104 Market data farm connection is OK:cashfarm\n", - "ERROR -1 2104 Market data farm connection is OK:usfarm\n", - "ERROR -1 2106 HMDS data farm connection is OK:ushmds\n", - "ERROR -1 2158 Sec-def data farm connection is OK:secdefnj\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Historical Data Ended\n", - "Time taken to pull data: 48.37 seconds\n", - " Date Open High Low Close Volume\n", - "0 20230502 18:25:00 66.59 66.59 66.59 66.59 1\n", - "1 20230502 18:30:00 66.59 66.59 66.59 66.59 0\n", - "2 20230502 18:35:00 66.59 66.59 66.59 66.59 0\n", - "3 20230502 18:40:00 66.59 66.59 66.59 66.59 0\n", - "4 20230502 18:45:00 66.59 66.59 66.59 66.59 0\n" - ] - } - ], - "source": [ - "## TRAINING DATA PULL (MAY - JULY 2024 - 5 MINS)\n", - "\n", - "from ibapi.client import EClient\n", - "from ibapi.wrapper import EWrapper\n", - "from ibapi.contract import Contract\n", - "import threading\n", - "import time\n", - "import pandas as pd\n", - "\n", - "class IBApi(EWrapper, EClient):\n", - " def __init__(self):\n", - " EClient.__init__(self, self)\n", - " self.data = [] # Store data\n", - " self.data_retrieved = False # Flag to check if data retrieval is complete\n", - "\n", - " def historicalData(self, reqId, bar):\n", - " self.data.append({\n", - " \"Date\": bar.date,\n", - " \"Open\": bar.open,\n", - " \"High\": bar.high,\n", - " \"Low\": bar.low,\n", - " \"Close\": bar.close,\n", - " \"Volume\": bar.volume\n", - " })\n", - "\n", - " def historicalDataEnd(self, reqId, start, end):\n", - " print(\"Historical Data Ended\")\n", - " self.df = pd.DataFrame(self.data)\n", - " self.data_retrieved = True # Set the flag to True to indicate data retrieval completion\n", - " self.disconnect()\n", - "\n", - "class IBApp:\n", - " def __init__(self):\n", - " self.app = IBApi()\n", - "\n", - " def connect(self):\n", - " self.app.connect(\"127.0.0.1\", 7496, 0)\n", - " thread = threading.Thread(target=self.run_app, daemon=True)\n", - " thread.start()\n", - " time.sleep(1)\n", - "\n", - " def run_app(self):\n", - " self.app.run()\n", - "\n", - " def request_training_data(self):\n", - " contract = Contract()\n", - " contract.symbol = \"CL\"\n", - " contract.secType = \"FUT\"\n", - " contract.exchange = \"NYMEX\"\n", - " contract.currency = \"USD\"\n", - " contract.lastTradeDateOrContractMonth = \"202412\" # November 2024 contract\n", - "\n", - " # Set parameters for data pull\n", - " end_date = \"20230730 23:59:59 UTC\" # Example end date in UTC\n", - " duration = \"3 M\" # 1 month duration\n", - " bar_size = \"5 mins\"\n", - "\n", - " # Record start time\n", - " start_time = time.time()\n", - "\n", - " # Request historical data\n", - " self.app.reqHistoricalData(\n", - " reqId=1,\n", - " contract=contract,\n", - " endDateTime=end_date,\n", - " durationStr=duration,\n", - " barSizeSetting=bar_size,\n", - " whatToShow='TRADES',\n", - " useRTH=0,\n", - " formatDate=1,\n", - " keepUpToDate=False,\n", - " chartOptions=[]\n", - " )\n", - "\n", - " # Wait until data retrieval is complete\n", - " while not self.app.data_retrieved:\n", - " time.sleep(0.1) # Small sleep interval to prevent busy-waiting\n", - "\n", - " # Record end time and calculate elapsed time\n", - " end_time = time.time()\n", - " elapsed_time = end_time - start_time\n", - " print(f\"Time taken to pull data: {elapsed_time:.2f} seconds\")\n", - "\n", - " def disconnect(self):\n", - " self.app.disconnect()\n", - "\n", - "# Instantiate and connect the app\n", - "app = IBApp()\n", - "app.connect()\n", - "\n", - "# Request training data\n", - "app.request_training_data()\n", - "\n", - "# Access the DataFrame\n", - "train_data = app.app.df if hasattr(app.app, 'df') else pd.DataFrame()\n", - "\n", - "# Disconnect from API\n", - "app.disconnect()\n", - "\n", - "# Display the training data\n", - "print(train_data.head())" - ] - }, - { - "cell_type": "code", - "execution_count": 16, - "id": "6768b1e6-3521-47d9-bb2d-1b5995227c5e", - "metadata": {}, - "outputs": [ - { - "data": { - "text/plain": [ - "17535" - ] - }, - "execution_count": 16, - "metadata": {}, - "output_type": "execute_result" - } - ], - "source": [ - "#train_data.to_csv(\"3_month_training_data.csv\")\n", - "len(train_data)" - ] - }, - { - "cell_type": "code", - "execution_count": 21, - "id": "cce534c5-6c19-4d4d-9a3a-e6a16c2ffe5c", - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "ERROR -1 2104 Market data farm connection is OK:usfarm.nj\n", - "ERROR -1 2104 Market data farm connection is OK:usfuture\n", - "ERROR -1 2104 Market data farm connection is OK:cashfarm\n", - "ERROR -1 2104 Market data farm connection is OK:usfarm\n", - "ERROR -1 2106 HMDS data farm connection is OK:ushmds\n", - "ERROR -1 2158 Sec-def data farm connection is OK:secdefnj\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Historical Data Ended\n", - "Time taken to pull data: 36.40 seconds\n", - " Date Open High Low Close Volume\n", - "0 20240804 18:00:00 71.99 72.30 71.99 72.21 95\n", - "1 20240804 18:05:00 72.15 72.15 71.86 71.86 214\n", - "2 20240804 18:10:00 71.88 71.94 71.85 71.94 65\n", - "3 20240804 18:15:00 71.85 71.87 71.63 71.77 63\n", - "4 20240804 18:20:00 71.75 71.75 71.70 71.73 66\n" - ] - } - ], - "source": [ - "## TESTING DATA (AUGUST - OCTOBER 2024 - 5 MINS)\n", - "\n", - "from ibapi.client import EClient\n", - "from ibapi.wrapper import EWrapper\n", - "from ibapi.contract import Contract\n", - "import threading\n", - "import time\n", - "import pandas as pd\n", - "\n", - "class IBApi(EWrapper, EClient):\n", - " def __init__(self):\n", - " EClient.__init__(self, self)\n", - " self.data = [] # Store data\n", - " self.data_retrieved = False # Flag to check if data retrieval is complete\n", - "\n", - " def historicalData(self, reqId, bar):\n", - " self.data.append({\n", - " \"Date\": bar.date,\n", - " \"Open\": bar.open,\n", - " \"High\": bar.high,\n", - " \"Low\": bar.low,\n", - " \"Close\": bar.close,\n", - " \"Volume\": bar.volume\n", - " })\n", - "\n", - " def historicalDataEnd(self, reqId, start, end):\n", - " print(\"Historical Data Ended\")\n", - " self.df = pd.DataFrame(self.data)\n", - " self.data_retrieved = True # Set the flag to True to indicate data retrieval completion\n", - " self.disconnect()\n", - "\n", - "class IBApp:\n", - " def __init__(self):\n", - " self.app = IBApi()\n", - "\n", - " def connect(self):\n", - " self.app.connect(\"127.0.0.1\", 7496, 0)\n", - " thread = threading.Thread(target=self.run_app, daemon=True)\n", - " thread.start()\n", - " time.sleep(1)\n", - "\n", - " def run_app(self):\n", - " self.app.run()\n", - "\n", - " def request_training_data(self):\n", - " contract = Contract()\n", - " contract.symbol = \"CL\"\n", - " contract.secType = \"FUT\"\n", - " contract.exchange = \"NYMEX\"\n", - " contract.currency = \"USD\"\n", - " contract.lastTradeDateOrContractMonth = \"202412\" # November 2024 contract\n", - "\n", - " # Set parameters for data pull\n", - " end_date = \"20241031 23:59:59 UTC\" # Example end date in UTC\n", - " duration = \"3 M\" # 1 month duration\n", - " bar_size = \"5 mins\"\n", - "\n", - " # Record start time\n", - " start_time = time.time()\n", - "\n", - " # Request historical data\n", - " self.app.reqHistoricalData(\n", - " reqId=1,\n", - " contract=contract,\n", - " endDateTime=end_date,\n", - " durationStr=duration,\n", - " barSizeSetting=bar_size,\n", - " whatToShow='TRADES',\n", - " useRTH=0,\n", - " formatDate=1,\n", - " keepUpToDate=False,\n", - " chartOptions=[]\n", - " )\n", - "\n", - " # Wait until data retrieval is complete\n", - " while not self.app.data_retrieved:\n", - " time.sleep(0.1) # Small sleep interval to prevent busy-waiting\n", - "\n", - " # Record end time and calculate elapsed time\n", - " end_time = time.time()\n", - " elapsed_time = end_time - start_time\n", - " print(f\"Time taken to pull data: {elapsed_time:.2f} seconds\")\n", - "\n", - " def disconnect(self):\n", - " self.app.disconnect()\n", - "\n", - "# Instantiate and connect the app\n", - "app = IBApp()\n", - "app.connect()\n", - "\n", - "# Request testing data\n", - "app.request_training_data()\n", - "\n", - "# Access the DataFrame\n", - "test_data = app.app.df if hasattr(app.app, 'df') else pd.DataFrame()\n", - "\n", - "# Disconnect from API\n", - "app.disconnect()\n", - "\n", - "# Display the testing data\n", - "print(test_data.head())" - ] - }, - { - "cell_type": "code", - "execution_count": 25, - "id": "ec3c6270-cedd-4579-a478-ac6f35d44694", - "metadata": {}, - "outputs": [], - "source": [ - "test_data.to_csv(\"3_month_testing_data.csv\")" - ] - }, - { - "cell_type": "code", - "execution_count": 33, - "id": "1b2404c3-b69a-489b-a884-295e2d739856", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mean Squared Error (MSE): 0.14\n", - "Mean Absolute Error (MAE): 0.14\n", - " Actual Predicted\n", - "0 72.21 72.13747\n", - "1 71.86 72.09486\n", - "2 71.94 71.88940\n", - "3 71.77 71.83409\n", - "4 71.73 71.71240\n" - ] - } - ], - "source": [ - "import pandas as pd\n", - "from sklearn.ensemble import RandomForestRegressor\n", - "from sklearn.metrics import mean_squared_error, mean_absolute_error\n", - "\n", - "# Load the training data\n", - "training_data = pd.read_csv(\"3_month_training_data.csv\")\n", - "testing_data = pd.read_csv(\"3_month_testing_data.csv\")\n", - "\n", - "# Preprocess the data: Drop unnecessary columns and separate features and target\n", - "# For both training and testing data, drop \"Unnamed: 0\" and \"Date\" columns\n", - "training_data = training_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "testing_data = testing_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "\n", - "# Split into features (X) and target (y) for both datasets\n", - "X_train = training_data.drop(columns=[\"Close\"])\n", - "y_train = training_data[\"Close\"]\n", - "X_test = testing_data.drop(columns=[\"Close\"])\n", - "y_test = testing_data[\"Close\"]\n", - "\n", - "# Train the Random Forest model\n", - "rf_model = RandomForestRegressor(n_estimators=1000, random_state=42)\n", - "rf_model.fit(X_train, y_train)\n", - "\n", - "# Make predictions on the testing data\n", - "y_pred = rf_model.predict(X_test)\n", - "\n", - "# Evaluate the model\n", - "mse = mean_squared_error(y_test, y_pred)\n", - "mae = mean_absolute_error(y_test, y_pred)\n", - "\n", - "# Print the evaluation metrics\n", - "print(f\"Mean Squared Error (MSE): {mse:.2f}\")\n", - "print(f\"Mean Absolute Error (MAE): {mae:.2f}\")\n", - "\n", - "# Optionally, view a few predictions\n", - "predictions = pd.DataFrame({\"Actual\": y_test, \"Predicted\": y_pred})\n", - "print(predictions.head())" - ] - }, - { - "cell_type": "code", - "execution_count": 35, - "id": "2c56076f-956f-4285-b14c-042496fa39eb", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Fitting 3 folds for each of 81 candidates, totalling 243 fits\n" - ] - } - ], - "source": [ - "from sklearn.model_selection import GridSearchCV\n", - "\n", - "param_grid = {\n", - " 'n_estimators': [100, 200, 300],\n", - " 'max_depth': [10, 20, None],\n", - " 'min_samples_split': [2, 5, 10],\n", - " 'min_samples_leaf': [1, 2, 4]\n", - "}\n", - "\n", - "grid_search = GridSearchCV(\n", - " estimator=RandomForestRegressor(random_state=42),\n", - " param_grid=param_grid,\n", - " scoring='neg_mean_squared_error',\n", - " cv=3,\n", - " n_jobs=-1,\n", - " verbose=2\n", - ")\n", - "\n", - "grid_search.fit(X_train, y_train)\n", - "best_model = grid_search.best_estimator_\n", - "\n", - "# Use best_model to make predictions\n", - "y_pred = best_model.predict(X_test)" - ] - }, - { - "cell_type": "code", - "execution_count": 49, - "id": "b9a15823-89b5-41ff-99d8-db18eea6d5a0", - "metadata": {}, - "outputs": [ - { - "ename": "ValueError", - "evalue": "The feature names should match those that were passed during fit.\nFeature names unseen at fit time:\n- Close_lag_1\n", - "output_type": "error", - "traceback": [ - "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m", - "\u001b[1;31mValueError\u001b[0m Traceback (most recent call last)", - "Cell \u001b[1;32mIn[49], line 2\u001b[0m\n\u001b[0;32m 1\u001b[0m \u001b[38;5;66;03m# Make predictions on the test set\u001b[39;00m\n\u001b[1;32m----> 2\u001b[0m y_pred \u001b[38;5;241m=\u001b[39m best_model\u001b[38;5;241m.\u001b[39mpredict(X_test)\n\u001b[0;32m 4\u001b[0m \u001b[38;5;66;03m# Evaluate performance with metrics like Mean Squared Error (MSE) and Mean Absolute Error (MAE)\u001b[39;00m\n\u001b[0;32m 5\u001b[0m mse \u001b[38;5;241m=\u001b[39m mean_squared_error(y_test, y_pred)\n", - "File \u001b[1;32m~\\anaconda3\\Lib\\site-packages\\sklearn\\ensemble\\_forest.py:1064\u001b[0m, in \u001b[0;36mForestRegressor.predict\u001b[1;34m(self, X)\u001b[0m\n\u001b[0;32m 1062\u001b[0m check_is_fitted(\u001b[38;5;28mself\u001b[39m)\n\u001b[0;32m 1063\u001b[0m \u001b[38;5;66;03m# Check data\u001b[39;00m\n\u001b[1;32m-> 1064\u001b[0m X \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_validate_X_predict(X)\n\u001b[0;32m 1066\u001b[0m \u001b[38;5;66;03m# Assign chunk of trees to jobs\u001b[39;00m\n\u001b[0;32m 1067\u001b[0m n_jobs, _, _ \u001b[38;5;241m=\u001b[39m _partition_estimators(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mn_estimators, \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mn_jobs)\n", - "File \u001b[1;32m~\\anaconda3\\Lib\\site-packages\\sklearn\\ensemble\\_forest.py:641\u001b[0m, in \u001b[0;36mBaseForest._validate_X_predict\u001b[1;34m(self, X)\u001b[0m\n\u001b[0;32m 638\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m 639\u001b[0m force_all_finite \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mTrue\u001b[39;00m\n\u001b[1;32m--> 641\u001b[0m X \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_validate_data(\n\u001b[0;32m 642\u001b[0m X,\n\u001b[0;32m 643\u001b[0m dtype\u001b[38;5;241m=\u001b[39mDTYPE,\n\u001b[0;32m 644\u001b[0m accept_sparse\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mcsr\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[0;32m 645\u001b[0m reset\u001b[38;5;241m=\u001b[39m\u001b[38;5;28;01mFalse\u001b[39;00m,\n\u001b[0;32m 646\u001b[0m force_all_finite\u001b[38;5;241m=\u001b[39mforce_all_finite,\n\u001b[0;32m 647\u001b[0m )\n\u001b[0;32m 648\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m issparse(X) \u001b[38;5;129;01mand\u001b[39;00m (X\u001b[38;5;241m.\u001b[39mindices\u001b[38;5;241m.\u001b[39mdtype \u001b[38;5;241m!=\u001b[39m np\u001b[38;5;241m.\u001b[39mintc \u001b[38;5;129;01mor\u001b[39;00m X\u001b[38;5;241m.\u001b[39mindptr\u001b[38;5;241m.\u001b[39mdtype \u001b[38;5;241m!=\u001b[39m np\u001b[38;5;241m.\u001b[39mintc):\n\u001b[0;32m 649\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mNo support for np.int64 index based sparse matrices\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n", - "File \u001b[1;32m~\\anaconda3\\Lib\\site-packages\\sklearn\\base.py:608\u001b[0m, in \u001b[0;36mBaseEstimator._validate_data\u001b[1;34m(self, X, y, reset, validate_separately, cast_to_ndarray, **check_params)\u001b[0m\n\u001b[0;32m 537\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_validate_data\u001b[39m(\n\u001b[0;32m 538\u001b[0m \u001b[38;5;28mself\u001b[39m,\n\u001b[0;32m 539\u001b[0m X\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mno_validation\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[1;32m (...)\u001b[0m\n\u001b[0;32m 544\u001b[0m \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mcheck_params,\n\u001b[0;32m 545\u001b[0m ):\n\u001b[0;32m 546\u001b[0m \u001b[38;5;250m \u001b[39m\u001b[38;5;124;03m\"\"\"Validate input data and set or check the `n_features_in_` attribute.\u001b[39;00m\n\u001b[0;32m 547\u001b[0m \n\u001b[0;32m 548\u001b[0m \u001b[38;5;124;03m Parameters\u001b[39;00m\n\u001b[1;32m (...)\u001b[0m\n\u001b[0;32m 606\u001b[0m \u001b[38;5;124;03m validated.\u001b[39;00m\n\u001b[0;32m 607\u001b[0m \u001b[38;5;124;03m \"\"\"\u001b[39;00m\n\u001b[1;32m--> 608\u001b[0m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_check_feature_names(X, reset\u001b[38;5;241m=\u001b[39mreset)\n\u001b[0;32m 610\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m y \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_get_tags()[\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrequires_y\u001b[39m\u001b[38;5;124m\"\u001b[39m]:\n\u001b[0;32m 611\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\n\u001b[0;32m 612\u001b[0m \u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mThis \u001b[39m\u001b[38;5;132;01m{\u001b[39;00m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__class__\u001b[39m\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__name__\u001b[39m\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m estimator \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m 613\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mrequires y to be passed, but the target y is None.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m 614\u001b[0m )\n", - "File \u001b[1;32m~\\anaconda3\\Lib\\site-packages\\sklearn\\base.py:535\u001b[0m, in \u001b[0;36mBaseEstimator._check_feature_names\u001b[1;34m(self, X, reset)\u001b[0m\n\u001b[0;32m 530\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m missing_names \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m unexpected_names:\n\u001b[0;32m 531\u001b[0m message \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m (\n\u001b[0;32m 532\u001b[0m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mFeature names must be in the same order as they were in fit.\u001b[39m\u001b[38;5;130;01m\\n\u001b[39;00m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m 533\u001b[0m )\n\u001b[1;32m--> 535\u001b[0m \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(message)\n", - "\u001b[1;31mValueError\u001b[0m: The feature names should match those that were passed during fit.\nFeature names unseen at fit time:\n- Close_lag_1\n" - ] - } - ], - "source": [ - "# Make predictions on the test set\n", - "y_pred = best_model.predict(X_test)\n", - "\n", - "# Evaluate performance with metrics like Mean Squared Error (MSE) and Mean Absolute Error (MAE)\n", - "mse = mean_squared_error(y_test, y_pred)\n", - "mae = mean_absolute_error(y_test, y_pred)\n", - "\n", - "print(f\"Best Model MSE: {mse:.2f}\")\n", - "print(f\"Best Model MAE: {mae:.2f}\")\n", - "print(\"Best Hyperparameters:\", grid_search.best_params_)" - ] - }, - { - "cell_type": "code", - "execution_count": 43, - "id": "fadf8772-c6d3-409c-b26b-bb90117048fd", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Mean Squared Error (MSE): 0.14\n", - "Mean Absolute Error (MAE): 0.14\n", - " Actual Predicted\n", - "1 71.86 72.12295\n", - "2 71.94 71.88483\n", - "3 71.77 71.83399\n", - "4 71.73 71.71186\n", - "5 71.79 71.79762\n" - ] - } - ], - "source": [ - "import pandas as pd\n", - "from sklearn.ensemble import RandomForestRegressor\n", - "from sklearn.metrics import mean_squared_error, mean_absolute_error\n", - "\n", - "# Create a lagged feature for Close price\n", - "training_data['Close_lag_1'] = training_data['Close'].shift(1)\n", - "training_data.dropna(inplace=True) # Drop rows with NaN values after shifting\n", - "\n", - "# Do the same for testing data\n", - "testing_data['Close_lag_1'] = testing_data['Close'].shift(1)\n", - "testing_data.dropna(inplace=True)\n", - "\n", - "# Update X_train and X_test with new features\n", - "X_train = training_data.drop(columns=[\"Close\"])\n", - "y_train = training_data[\"Close\"]\n", - "X_test = testing_data.drop(columns=[\"Close\"])\n", - "y_test = testing_data[\"Close\"]\n", - "\n", - "# Train the Random Forest model\n", - "rf_model = RandomForestRegressor(n_estimators=1000, random_state=42)\n", - "rf_model.fit(X_train, y_train)\n", - "\n", - "# Make predictions on the testing data\n", - "y_pred = rf_model.predict(X_test)\n", - "\n", - "# Evaluate the model\n", - "mse = mean_squared_error(y_test, y_pred)\n", - "mae = mean_absolute_error(y_test, y_pred)\n", - "\n", - "# Print the evaluation metrics\n", - "print(f\"Mean Squared Error (MSE): {mse:.2f}\")\n", - "print(f\"Mean Absolute Error (MAE): {mae:.2f}\")\n", - "\n", - "# Optionally, view a few predictions\n", - "predictions = pd.DataFrame({\"Actual\": y_test, \"Predicted\": y_pred})\n", - "print(predictions.head())" - ] - }, - { - "cell_type": "code", - "execution_count": 47, - "id": "e440ae19-0407-403d-91f7-b8eca4684e1c", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Collecting xgboost\n", - " Downloading xgboost-2.1.2-py3-none-win_amd64.whl.metadata (2.1 kB)\n", - "Requirement already satisfied: numpy in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from xgboost) (1.26.4)\n", - "Requirement already satisfied: scipy in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from xgboost) (1.13.1)\n", - "Downloading xgboost-2.1.2-py3-none-win_amd64.whl (124.9 MB)\n", - " ---------------------------------------- 0.0/124.9 MB ? eta -:--:--\n", - " --------------------------------------- 2.1/124.9 MB 11.8 MB/s eta 0:00:11\n", - " - -------------------------------------- 4.2/124.9 MB 10.1 MB/s eta 0:00:13\n", - " -- ------------------------------------- 6.6/124.9 MB 10.6 MB/s eta 0:00:12\n", - " -- ------------------------------------- 8.4/124.9 MB 10.0 MB/s eta 0:00:12\n", - " --- ------------------------------------ 10.7/124.9 MB 10.0 MB/s eta 0:00:12\n", - " ---- ----------------------------------- 13.1/124.9 MB 10.1 MB/s eta 0:00:12\n", - " ---- ----------------------------------- 15.5/124.9 MB 10.4 MB/s eta 0:00:11\n", - " ----- ---------------------------------- 17.8/124.9 MB 10.5 MB/s eta 0:00:11\n", - " ------ --------------------------------- 19.9/124.9 MB 10.5 MB/s eta 0:00:11\n", - " ------- -------------------------------- 22.0/124.9 MB 10.5 MB/s eta 0:00:10\n", - " ------- -------------------------------- 24.4/124.9 MB 10.4 MB/s eta 0:00:10\n", - " -------- ------------------------------- 26.7/124.9 MB 10.5 MB/s eta 0:00:10\n", - " --------- ------------------------------ 28.8/124.9 MB 10.6 MB/s eta 0:00:10\n", - " ---------- ----------------------------- 31.5/124.9 MB 10.6 MB/s eta 0:00:09\n", - " ---------- ----------------------------- 33.8/124.9 MB 10.6 MB/s eta 0:00:09\n", - " ----------- ---------------------------- 36.2/124.9 MB 10.7 MB/s eta 0:00:09\n", - " ------------ --------------------------- 38.5/124.9 MB 10.7 MB/s eta 0:00:09\n", - " ------------- -------------------------- 40.6/124.9 MB 10.7 MB/s eta 0:00:08\n", - " ------------- -------------------------- 43.0/124.9 MB 10.6 MB/s eta 0:00:08\n", - " -------------- ------------------------- 45.1/124.9 MB 10.7 MB/s eta 0:00:08\n", - " --------------- ------------------------ 47.4/124.9 MB 10.7 MB/s eta 0:00:08\n", - " --------------- ------------------------ 49.8/124.9 MB 10.6 MB/s eta 0:00:08\n", - " ---------------- ----------------------- 52.4/124.9 MB 10.7 MB/s eta 0:00:07\n", - " ----------------- ---------------------- 54.8/124.9 MB 10.7 MB/s eta 0:00:07\n", - " ------------------ --------------------- 57.1/124.9 MB 10.7 MB/s eta 0:00:07\n", - " ------------------ --------------------- 59.2/124.9 MB 10.7 MB/s eta 0:00:07\n", - " ------------------- -------------------- 61.9/124.9 MB 10.7 MB/s eta 0:00:06\n", - " -------------------- ------------------- 64.2/124.9 MB 10.8 MB/s eta 0:00:06\n", - " --------------------- ------------------ 66.3/124.9 MB 10.7 MB/s eta 0:00:06\n", - " --------------------- ------------------ 68.7/124.9 MB 10.8 MB/s eta 0:00:06\n", - " ---------------------- ----------------- 71.0/124.9 MB 10.8 MB/s eta 0:00:06\n", - " ----------------------- ---------------- 73.4/124.9 MB 10.8 MB/s eta 0:00:05\n", - " ------------------------ --------------- 75.8/124.9 MB 10.8 MB/s eta 0:00:05\n", - " ------------------------- -------------- 78.1/124.9 MB 10.8 MB/s eta 0:00:05\n", - " ------------------------- -------------- 80.7/124.9 MB 10.8 MB/s eta 0:00:05\n", - " -------------------------- ------------- 83.1/124.9 MB 10.8 MB/s eta 0:00:04\n", - " --------------------------- ------------ 85.5/124.9 MB 10.8 MB/s eta 0:00:04\n", - " ---------------------------- ----------- 87.8/124.9 MB 10.8 MB/s eta 0:00:04\n", - " ---------------------------- ----------- 90.2/124.9 MB 10.8 MB/s eta 0:00:04\n", - " ----------------------------- ---------- 92.5/124.9 MB 10.8 MB/s eta 0:00:03\n", - " ------------------------------ --------- 95.2/124.9 MB 10.8 MB/s eta 0:00:03\n", - " ------------------------------- -------- 97.8/124.9 MB 10.9 MB/s eta 0:00:03\n", - " -------------------------------- ------ 103.8/124.9 MB 11.3 MB/s eta 0:00:02\n", - " ----------------------------------- --- 112.2/124.9 MB 11.9 MB/s eta 0:00:02\n", - " -------------------------------------- 121.9/124.9 MB 12.7 MB/s eta 0:00:01\n", - " -------------------------------------- 124.8/124.9 MB 12.9 MB/s eta 0:00:01\n", - " --------------------------------------- 124.9/124.9 MB 12.6 MB/s eta 0:00:00\n", - "Installing collected packages: xgboost\n", - "Successfully installed xgboost-2.1.2\n", - "XGBoost MSE: 0.16, MAE: 0.17\n" - ] - } - ], - "source": [ - "!pip install xgboost\n", - "from xgboost import XGBRegressor\n", - "\n", - "xgb_model = XGBRegressor(n_estimators=100, max_depth=5, learning_rate=0.1, random_state=42)\n", - "xgb_model.fit(X_train, y_train)\n", - "y_pred = xgb_model.predict(X_test)\n", - "\n", - "# Evaluate the XGBoost model\n", - "mse = mean_squared_error(y_test, y_pred)\n", - "mae = mean_absolute_error(y_test, y_pred)\n", - "print(f\"XGBoost MSE: {mse:.2f}, MAE: {mae:.2f}\")" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "id": "ec41eca3-ac8f-427a-8cd4-df783940fa40", - "metadata": {}, - "outputs": [], - "source": [ - "## As we can see this training model is not as good as RandomForest" - ] - }, - { - "cell_type": "code", - "execution_count": 51, - "id": "63ac0e5c-6590-4541-b618-cf60c3af0f00", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Collecting tensorflow\n", - " Downloading tensorflow-2.18.0-cp312-cp312-win_amd64.whl.metadata (3.3 kB)\n", - "Collecting tensorflow-intel==2.18.0 (from tensorflow)\n", - " Downloading tensorflow_intel-2.18.0-cp312-cp312-win_amd64.whl.metadata (4.9 kB)\n", - "Collecting absl-py>=1.0.0 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading absl_py-2.1.0-py3-none-any.whl.metadata (2.3 kB)\n", - "Collecting astunparse>=1.6.0 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading astunparse-1.6.3-py2.py3-none-any.whl.metadata (4.4 kB)\n", - "Collecting flatbuffers>=24.3.25 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading flatbuffers-24.3.25-py2.py3-none-any.whl.metadata (850 bytes)\n", - "Collecting gast!=0.5.0,!=0.5.1,!=0.5.2,>=0.2.1 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading gast-0.6.0-py3-none-any.whl.metadata (1.3 kB)\n", - "Collecting google-pasta>=0.1.1 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading google_pasta-0.2.0-py3-none-any.whl.metadata (814 bytes)\n", - "Collecting libclang>=13.0.0 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading libclang-18.1.1-py2.py3-none-win_amd64.whl.metadata (5.3 kB)\n", - "Collecting opt-einsum>=2.3.2 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading opt_einsum-3.4.0-py3-none-any.whl.metadata (6.3 kB)\n", - "Requirement already satisfied: packaging in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (23.2)\n", - "Requirement already satisfied: protobuf!=4.21.0,!=4.21.1,!=4.21.2,!=4.21.3,!=4.21.4,!=4.21.5,<6.0.0dev,>=3.20.3 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (3.20.3)\n", - "Requirement already satisfied: requests<3,>=2.21.0 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (2.32.2)\n", - "Requirement already satisfied: setuptools in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (69.5.1)\n", - "Requirement already satisfied: six>=1.12.0 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (1.16.0)\n", - "Collecting termcolor>=1.1.0 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading termcolor-2.5.0-py3-none-any.whl.metadata (6.1 kB)\n", - "Requirement already satisfied: typing-extensions>=3.6.6 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (4.11.0)\n", - "Requirement already satisfied: wrapt>=1.11.0 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (1.14.1)\n", - "Collecting grpcio<2.0,>=1.24.3 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading grpcio-1.67.1-cp312-cp312-win_amd64.whl.metadata (4.0 kB)\n", - "Collecting tensorboard<2.19,>=2.18 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading tensorboard-2.18.0-py3-none-any.whl.metadata (1.6 kB)\n", - "Collecting keras>=3.5.0 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading keras-3.6.0-py3-none-any.whl.metadata (5.8 kB)\n", - "Requirement already satisfied: numpy<2.1.0,>=1.26.0 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (1.26.4)\n", - "Requirement already satisfied: h5py>=3.11.0 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorflow-intel==2.18.0->tensorflow) (3.11.0)\n", - "Collecting ml-dtypes<0.5.0,>=0.4.0 (from tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading ml_dtypes-0.4.1-cp312-cp312-win_amd64.whl.metadata (20 kB)\n", - "Requirement already satisfied: wheel<1.0,>=0.23.0 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from astunparse>=1.6.0->tensorflow-intel==2.18.0->tensorflow) (0.43.0)\n", - "Requirement already satisfied: rich in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from keras>=3.5.0->tensorflow-intel==2.18.0->tensorflow) (13.3.5)\n", - "Collecting namex (from keras>=3.5.0->tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading namex-0.0.8-py3-none-any.whl.metadata (246 bytes)\n", - "Collecting optree (from keras>=3.5.0->tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading optree-0.13.0-cp312-cp312-win_amd64.whl.metadata (48 kB)\n", - "Requirement already satisfied: charset-normalizer<4,>=2 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from requests<3,>=2.21.0->tensorflow-intel==2.18.0->tensorflow) (2.0.4)\n", - "Requirement already satisfied: idna<4,>=2.5 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from requests<3,>=2.21.0->tensorflow-intel==2.18.0->tensorflow) (3.7)\n", - "Requirement already satisfied: urllib3<3,>=1.21.1 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from requests<3,>=2.21.0->tensorflow-intel==2.18.0->tensorflow) (2.2.2)\n", - "Requirement already satisfied: certifi>=2017.4.17 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from requests<3,>=2.21.0->tensorflow-intel==2.18.0->tensorflow) (2024.8.30)\n", - "Requirement already satisfied: markdown>=2.6.8 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorboard<2.19,>=2.18->tensorflow-intel==2.18.0->tensorflow) (3.4.1)\n", - "Collecting tensorboard-data-server<0.8.0,>=0.7.0 (from tensorboard<2.19,>=2.18->tensorflow-intel==2.18.0->tensorflow)\n", - " Downloading tensorboard_data_server-0.7.2-py3-none-any.whl.metadata (1.1 kB)\n", - "Requirement already satisfied: werkzeug>=1.0.1 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from tensorboard<2.19,>=2.18->tensorflow-intel==2.18.0->tensorflow) (3.0.3)\n", - "Requirement already satisfied: MarkupSafe>=2.1.1 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from werkzeug>=1.0.1->tensorboard<2.19,>=2.18->tensorflow-intel==2.18.0->tensorflow) (2.1.3)\n", - "Requirement already satisfied: markdown-it-py<3.0.0,>=2.2.0 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from rich->keras>=3.5.0->tensorflow-intel==2.18.0->tensorflow) (2.2.0)\n", - "Requirement already satisfied: pygments<3.0.0,>=2.13.0 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from rich->keras>=3.5.0->tensorflow-intel==2.18.0->tensorflow) (2.15.1)\n", - "Requirement already satisfied: mdurl~=0.1 in c:\\users\\gwitt\\anaconda3\\lib\\site-packages (from markdown-it-py<3.0.0,>=2.2.0->rich->keras>=3.5.0->tensorflow-intel==2.18.0->tensorflow) (0.1.0)\n", - "Downloading tensorflow-2.18.0-cp312-cp312-win_amd64.whl (7.5 kB)\n", - "Downloading tensorflow_intel-2.18.0-cp312-cp312-win_amd64.whl (390.3 MB)\n", - " ---------------------------------------- 0.0/390.3 MB ? eta -:--:--\n", - " ---------------------------------------- 2.1/390.3 MB 11.7 MB/s eta 0:00:34\n", - " ---------------------------------------- 4.5/390.3 MB 11.2 MB/s eta 0:00:35\n", - " --------------------------------------- 7.1/390.3 MB 11.5 MB/s eta 0:00:34\n", - " --------------------------------------- 9.4/390.3 MB 11.3 MB/s eta 0:00:34\n", - " - -------------------------------------- 11.5/390.3 MB 11.1 MB/s eta 0:00:35\n", - " - -------------------------------------- 14.2/390.3 MB 11.1 MB/s eta 0:00:34\n", - " - -------------------------------------- 16.5/390.3 MB 11.2 MB/s eta 0:00:34\n", - " - -------------------------------------- 18.9/390.3 MB 11.1 MB/s eta 0:00:34\n", - " -- ------------------------------------- 21.2/390.3 MB 11.2 MB/s eta 0:00:33\n", - " -- ------------------------------------- 23.6/390.3 MB 11.1 MB/s eta 0:00:33\n", - " -- ------------------------------------- 26.0/390.3 MB 11.0 MB/s eta 0:00:34\n", - " -- ------------------------------------- 28.3/390.3 MB 11.1 MB/s eta 0:00:33\n", - " --- ------------------------------------ 30.7/390.3 MB 11.1 MB/s eta 0:00:33\n", - " --- ------------------------------------ 32.8/390.3 MB 11.1 MB/s eta 0:00:33\n", - " --- ------------------------------------ 35.1/390.3 MB 11.0 MB/s eta 0:00:33\n", - " --- ------------------------------------ 37.7/390.3 MB 11.1 MB/s eta 0:00:32\n", - " ---- ----------------------------------- 39.6/390.3 MB 10.9 MB/s eta 0:00:33\n", - " ---- ----------------------------------- 41.7/390.3 MB 10.9 MB/s eta 0:00:33\n", - " ---- ----------------------------------- 44.0/390.3 MB 10.9 MB/s eta 0:00:32\n", - " ---- ----------------------------------- 46.7/390.3 MB 10.9 MB/s eta 0:00:32\n", - " ---- ----------------------------------- 48.8/390.3 MB 10.9 MB/s eta 0:00:32\n", - " ----- ---------------------------------- 51.1/390.3 MB 10.9 MB/s eta 0:00:32\n", - " ----- ---------------------------------- 53.0/390.3 MB 10.8 MB/s eta 0:00:32\n", - " ----- ---------------------------------- 55.6/390.3 MB 10.8 MB/s eta 0:00:32\n", - " ----- ---------------------------------- 57.9/390.3 MB 10.8 MB/s eta 0:00:31\n", - " ------ --------------------------------- 60.3/390.3 MB 10.9 MB/s eta 0:00:31\n", - " ------ --------------------------------- 62.7/390.3 MB 10.9 MB/s eta 0:00:31\n", - " ------ --------------------------------- 65.0/390.3 MB 10.8 MB/s eta 0:00:30\n", - " ------ --------------------------------- 67.4/390.3 MB 10.8 MB/s eta 0:00:30\n", - " ------- -------------------------------- 69.7/390.3 MB 10.8 MB/s eta 0:00:30\n", - " ------- -------------------------------- 72.1/390.3 MB 10.9 MB/s eta 0:00:30\n", - " ------- -------------------------------- 74.4/390.3 MB 10.9 MB/s eta 0:00:30\n", - " ------- -------------------------------- 76.8/390.3 MB 10.9 MB/s eta 0:00:29\n", - " -------- ------------------------------- 79.2/390.3 MB 10.9 MB/s eta 0:00:29\n", - " -------- ------------------------------- 81.5/390.3 MB 10.9 MB/s eta 0:00:29\n", - " -------- ------------------------------- 83.9/390.3 MB 10.9 MB/s eta 0:00:29\n", - " -------- ------------------------------- 86.5/390.3 MB 10.9 MB/s eta 0:00:28\n", - " --------- ------------------------------ 88.9/390.3 MB 10.9 MB/s eta 0:00:28\n", - " --------- ------------------------------ 91.2/390.3 MB 10.9 MB/s eta 0:00:28\n", - " --------- ------------------------------ 93.6/390.3 MB 10.9 MB/s eta 0:00:28\n", - " --------- ------------------------------ 95.9/390.3 MB 10.9 MB/s eta 0:00:28\n", - " ---------- ----------------------------- 98.0/390.3 MB 10.9 MB/s eta 0:00:27\n", - " ---------- ---------------------------- 100.7/390.3 MB 10.9 MB/s eta 0:00:27\n", - " ---------- ---------------------------- 103.3/390.3 MB 10.9 MB/s eta 0:00:27\n", - " ---------- ---------------------------- 105.6/390.3 MB 10.9 MB/s eta 0:00:27\n", - " ---------- ---------------------------- 108.0/390.3 MB 10.9 MB/s eta 0:00:26\n", - " ----------- --------------------------- 110.1/390.3 MB 10.9 MB/s eta 0:00:26\n", - " ----------- --------------------------- 112.5/390.3 MB 10.9 MB/s eta 0:00:26\n", - " ----------- --------------------------- 114.8/390.3 MB 10.9 MB/s eta 0:00:26\n", - " ----------- --------------------------- 117.2/390.3 MB 10.9 MB/s eta 0:00:25\n", - " ----------- --------------------------- 119.5/390.3 MB 10.9 MB/s eta 0:00:25\n", - " ------------ -------------------------- 121.9/390.3 MB 10.9 MB/s eta 0:00:25\n", - " ------------ -------------------------- 124.3/390.3 MB 10.9 MB/s eta 0:00:25\n", - " ------------ -------------------------- 126.6/390.3 MB 10.9 MB/s eta 0:00:25\n", - " ------------ -------------------------- 129.0/390.3 MB 10.9 MB/s eta 0:00:24\n", - " ------------- ------------------------- 131.3/390.3 MB 10.9 MB/s eta 0:00:24\n", - " ------------- ------------------------- 133.4/390.3 MB 10.9 MB/s eta 0:00:24\n", - " ------------- ------------------------- 135.5/390.3 MB 10.9 MB/s eta 0:00:24\n", - " ------------- ------------------------- 137.9/390.3 MB 10.9 MB/s eta 0:00:24\n", - " -------------- ------------------------ 140.2/390.3 MB 10.9 MB/s eta 0:00:23\n", - " -------------- ------------------------ 142.9/390.3 MB 10.9 MB/s eta 0:00:23\n", - " -------------- ------------------------ 145.2/390.3 MB 10.9 MB/s eta 0:00:23\n", - " -------------- ------------------------ 147.8/390.3 MB 10.9 MB/s eta 0:00:23\n", - " -------------- ------------------------ 149.9/390.3 MB 10.9 MB/s eta 0:00:23\n", - " --------------- ----------------------- 152.6/390.3 MB 10.9 MB/s eta 0:00:22\n", - " --------------- ----------------------- 154.9/390.3 MB 10.9 MB/s eta 0:00:22\n", - " --------------- ----------------------- 157.5/390.3 MB 10.9 MB/s eta 0:00:22\n", - " --------------- ----------------------- 159.9/390.3 MB 11.0 MB/s eta 0:00:22\n", - " ---------------- ---------------------- 162.5/390.3 MB 11.0 MB/s eta 0:00:21\n", - " ---------------- ---------------------- 165.2/390.3 MB 11.0 MB/s eta 0:00:21\n", - " ---------------- ---------------------- 167.5/390.3 MB 11.0 MB/s eta 0:00:21\n", - " ---------------- ---------------------- 169.6/390.3 MB 11.0 MB/s eta 0:00:21\n", - " ----------------- --------------------- 172.2/390.3 MB 11.0 MB/s eta 0:00:20\n", - " ----------------- --------------------- 174.6/390.3 MB 11.0 MB/s eta 0:00:20\n", - " ----------------- --------------------- 177.2/390.3 MB 11.0 MB/s eta 0:00:20\n", - " ----------------- --------------------- 179.8/390.3 MB 11.0 MB/s eta 0:00:20\n", - " ------------------ -------------------- 184.3/390.3 MB 11.1 MB/s eta 0:00:19\n", - " ------------------- ------------------- 193.2/390.3 MB 11.5 MB/s eta 0:00:18\n", - " -------------------- ------------------ 201.9/390.3 MB 11.9 MB/s eta 0:00:16\n", - " --------------------- ----------------- 210.5/390.3 MB 12.3 MB/s eta 0:00:15\n", - " --------------------- ----------------- 218.9/390.3 MB 12.6 MB/s eta 0:00:14\n", - " ---------------------- ---------------- 221.5/390.3 MB 12.6 MB/s eta 0:00:14\n", - " ---------------------- ---------------- 224.7/390.3 MB 12.6 MB/s eta 0:00:14\n", - " ---------------------- ---------------- 227.3/390.3 MB 12.6 MB/s eta 0:00:13\n", - " ---------------------- ---------------- 229.6/390.3 MB 12.6 MB/s eta 0:00:13\n", - " ----------------------- --------------- 232.0/390.3 MB 12.6 MB/s eta 0:00:13\n", - " ----------------------- --------------- 234.6/390.3 MB 12.6 MB/s eta 0:00:13\n", - " ----------------------- --------------- 237.5/390.3 MB 12.6 MB/s eta 0:00:13\n", - " ----------------------- --------------- 239.9/390.3 MB 12.5 MB/s eta 0:00:12\n", - " ------------------------ -------------- 242.2/390.3 MB 12.5 MB/s eta 0:00:12\n", - " ------------------------ -------------- 247.2/390.3 MB 12.6 MB/s eta 0:00:12\n", - " ------------------------- ------------- 256.1/390.3 MB 13.0 MB/s eta 0:00:11\n", - " -------------------------- ------------ 264.2/390.3 MB 13.2 MB/s eta 0:00:10\n", - " --------------------------- ----------- 272.4/390.3 MB 13.6 MB/s eta 0:00:09\n", - " --------------------------- ----------- 278.7/390.3 MB 13.9 MB/s eta 0:00:09\n", - " ---------------------------- ---------- 282.1/390.3 MB 13.9 MB/s eta 0:00:08\n", - " ---------------------------- ---------- 284.7/390.3 MB 13.9 MB/s eta 0:00:08\n", - " ---------------------------- ---------- 287.0/390.3 MB 14.0 MB/s eta 0:00:08\n", - " ---------------------------- ---------- 289.7/390.3 MB 14.0 MB/s eta 0:00:08\n", - " ----------------------------- --------- 293.3/390.3 MB 14.0 MB/s eta 0:00:07\n", - " ------------------------------ -------- 302.3/390.3 MB 14.6 MB/s eta 0:00:07\n", - " ------------------------------- ------- 311.2/390.3 MB 15.1 MB/s eta 0:00:06\n", - " ------------------------------- ------- 319.8/390.3 MB 15.6 MB/s eta 0:00:05\n", - " -------------------------------- ------ 329.0/390.3 MB 16.2 MB/s eta 0:00:04\n", - " --------------------------------- ----- 337.6/390.3 MB 16.8 MB/s eta 0:00:04\n", - " ---------------------------------- ---- 346.0/390.3 MB 17.4 MB/s eta 0:00:03\n", - " ----------------------------------- --- 353.6/390.3 MB 18.0 MB/s eta 0:00:03\n", - " ------------------------------------ -- 362.8/390.3 MB 18.7 MB/s eta 0:00:02\n", - " ------------------------------------- - 370.7/390.3 MB 19.4 MB/s eta 0:00:02\n", - " ------------------------------------- - 379.1/390.3 MB 20.3 MB/s eta 0:00:01\n", - " -------------------------------------- 387.4/390.3 MB 21.2 MB/s eta 0:00:01\n", - " -------------------------------------- 390.1/390.3 MB 21.5 MB/s eta 0:00:01\n", - " -------------------------------------- 390.1/390.3 MB 21.5 MB/s eta 0:00:01\n", - " -------------------------------------- 390.1/390.3 MB 21.5 MB/s eta 0:00:01\n", - " -------------------------------------- 390.1/390.3 MB 21.5 MB/s eta 0:00:01\n", - " --------------------------------------- 390.3/390.3 MB 19.9 MB/s eta 0:00:00\n", - "Downloading absl_py-2.1.0-py3-none-any.whl (133 kB)\n", - "Downloading astunparse-1.6.3-py2.py3-none-any.whl (12 kB)\n", - "Downloading flatbuffers-24.3.25-py2.py3-none-any.whl (26 kB)\n", - "Downloading gast-0.6.0-py3-none-any.whl (21 kB)\n", - "Downloading google_pasta-0.2.0-py3-none-any.whl (57 kB)\n", - "Downloading grpcio-1.67.1-cp312-cp312-win_amd64.whl (4.3 MB)\n", - " ---------------------------------------- 0.0/4.3 MB ? eta -:--:--\n", - " --------------------- ------------------ 2.4/4.3 MB 11.2 MB/s eta 0:00:01\n", - " ---------------------------------------- 4.3/4.3 MB 10.9 MB/s eta 0:00:00\n", - "Downloading keras-3.6.0-py3-none-any.whl (1.2 MB)\n", - " ---------------------------------------- 0.0/1.2 MB ? eta -:--:--\n", - " ---------------------------------------- 1.2/1.2 MB 7.4 MB/s eta 0:00:00\n", - "Downloading libclang-18.1.1-py2.py3-none-win_amd64.whl (26.4 MB)\n", - " ---------------------------------------- 0.0/26.4 MB ? eta -:--:--\n", - " --- ------------------------------------ 2.6/26.4 MB 12.5 MB/s eta 0:00:02\n", - " ------- -------------------------------- 5.2/26.4 MB 12.2 MB/s eta 0:00:02\n", - " ----------- ---------------------------- 7.6/26.4 MB 12.0 MB/s eta 0:00:02\n", - " --------------- ------------------------ 10.5/26.4 MB 12.1 MB/s eta 0:00:02\n", - " ------------------- -------------------- 12.8/26.4 MB 12.0 MB/s eta 0:00:02\n", - " ----------------------- ---------------- 15.7/26.4 MB 12.1 MB/s eta 0:00:01\n", - " --------------------------- ------------ 18.4/26.4 MB 12.2 MB/s eta 0:00:01\n", - " ------------------------------- -------- 20.7/26.4 MB 12.1 MB/s eta 0:00:01\n", - " -------------------------------------- - 25.2/26.4 MB 13.0 MB/s eta 0:00:01\n", - " ---------------------------------------- 26.4/26.4 MB 12.9 MB/s eta 0:00:00\n", - "Downloading ml_dtypes-0.4.1-cp312-cp312-win_amd64.whl (127 kB)\n", - "Downloading opt_einsum-3.4.0-py3-none-any.whl (71 kB)\n", - "Downloading tensorboard-2.18.0-py3-none-any.whl (5.5 MB)\n", - " ---------------------------------------- 0.0/5.5 MB ? eta -:--:--\n", - " ---------------------------------------- 5.5/5.5 MB 33.6 MB/s eta 0:00:00\n", - "Downloading termcolor-2.5.0-py3-none-any.whl (7.8 kB)\n", - "Downloading tensorboard_data_server-0.7.2-py3-none-any.whl (2.4 kB)\n", - "Downloading namex-0.0.8-py3-none-any.whl (5.8 kB)\n", - "Downloading optree-0.13.0-cp312-cp312-win_amd64.whl (283 kB)\n", - "Installing collected packages: namex, libclang, flatbuffers, termcolor, tensorboard-data-server, optree, opt-einsum, ml-dtypes, grpcio, google-pasta, gast, astunparse, absl-py, tensorboard, keras, tensorflow-intel, tensorflow\n", - "Successfully installed absl-py-2.1.0 astunparse-1.6.3 flatbuffers-24.3.25 gast-0.6.0 google-pasta-0.2.0 grpcio-1.67.1 keras-3.6.0 libclang-18.1.1 ml-dtypes-0.4.1 namex-0.0.8 opt-einsum-3.4.0 optree-0.13.0 tensorboard-2.18.0 tensorboard-data-server-0.7.2 tensorflow-2.18.0 tensorflow-intel-2.18.0 termcolor-2.5.0\n" - ] - } - ], - "source": [ - "#!pip install tensorflow" - ] - }, - { - "cell_type": "code", - "execution_count": 57, - "id": "8898cfc8-99bb-42d0-894f-a26f000ec5f3", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 1/100\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "C:\\Users\\gwitt\\anaconda3\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n", - " super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 3ms/step - loss: 3027.9695 - mae: 50.1166 - val_loss: 580.0393 - val_mae: 17.1352\n", - "Epoch 2/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 33.9868 - mae: 3.1780 - val_loss: 549.8928 - val_mae: 17.4320\n", - "Epoch 3/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 8.0077 - mae: 1.4031 - val_loss: 417.2844 - val_mae: 15.0162\n", - "Epoch 4/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.8638 - mae: 0.6774 - val_loss: 341.3333 - val_mae: 13.4354\n", - "Epoch 5/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.4493 - mae: 0.3370 - val_loss: 294.7268 - val_mae: 12.3450\n", - "Epoch 6/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.1663 - mae: 0.2110 - val_loss: 263.7179 - val_mae: 11.5672\n", - "Epoch 7/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.1182 - mae: 0.1657 - val_loss: 244.0993 - val_mae: 11.0653\n", - "Epoch 8/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0743 - mae: 0.1364 - val_loss: 227.4102 - val_mae: 10.6388\n", - "Epoch 9/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0749 - mae: 0.1279 - val_loss: 215.9571 - val_mae: 10.3491\n", - "Epoch 10/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0351 - mae: 0.1073 - val_loss: 202.6387 - val_mae: 9.9880\n", - "Epoch 11/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0219 - mae: 0.0880 - val_loss: 187.2791 - val_mae: 9.5498\n", - "Epoch 12/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0170 - mae: 0.0771 - val_loss: 180.6176 - val_mae: 9.3664\n", - "Epoch 13/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0111 - mae: 0.0594 - val_loss: 167.2584 - val_mae: 8.9546\n", - "Epoch 14/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0160 - mae: 0.0626 - val_loss: 160.5615 - val_mae: 8.7712\n", - "Epoch 15/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0073 - mae: 0.0464 - val_loss: 151.2990 - val_mae: 8.4988\n", - "Epoch 16/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0097 - mae: 0.0535 - val_loss: 144.9969 - val_mae: 8.3089\n", - "Epoch 17/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0049 - mae: 0.0380 - val_loss: 144.3898 - val_mae: 8.3215\n", - "Epoch 18/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0030 - mae: 0.0343 - val_loss: 139.0735 - val_mae: 8.1323\n", - "Epoch 19/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0021 - mae: 0.0315 - val_loss: 136.6971 - val_mae: 8.0569\n", - "Epoch 20/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0021 - mae: 0.0307 - val_loss: 134.3677 - val_mae: 7.9754\n", - "Epoch 21/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0047 - mae: 0.0415 - val_loss: 133.5814 - val_mae: 7.9533\n", - "Epoch 22/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0026 - mae: 0.0342 - val_loss: 133.4585 - val_mae: 7.9509\n", - "Epoch 23/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0165 - mae: 0.0624 - val_loss: 126.6365 - val_mae: 7.7415\n", - "Epoch 24/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0154 - mae: 0.0773 - val_loss: 118.9955 - val_mae: 7.4735\n", - "Epoch 25/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0039 - mae: 0.0431 - val_loss: 114.6378 - val_mae: 7.3436\n", - "Epoch 26/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0137 - mae: 0.0578 - val_loss: 105.9814 - val_mae: 7.0309\n", - "Epoch 27/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0199 - mae: 0.0902 - val_loss: 109.9631 - val_mae: 7.1765\n", - "Epoch 28/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0030 - mae: 0.0348 - val_loss: 108.8627 - val_mae: 7.1402\n", - "Epoch 29/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0046 - mae: 0.0392 - val_loss: 111.1780 - val_mae: 7.2641\n", - "Epoch 30/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0144 - mae: 0.0695 - val_loss: 108.0607 - val_mae: 7.1151\n", - "Epoch 31/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0020 - mae: 0.0314 - val_loss: 107.1755 - val_mae: 7.0774\n", - "Epoch 32/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0038 - mae: 0.0424 - val_loss: 108.9577 - val_mae: 7.1964\n", - "Epoch 33/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0125 - mae: 0.0731 - val_loss: 104.9709 - val_mae: 6.9971\n", - "Epoch 34/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0091 - mae: 0.0606 - val_loss: 102.2909 - val_mae: 6.8926\n", - "Epoch 35/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0086 - mae: 0.0583 - val_loss: 103.8827 - val_mae: 6.9738\n", - "Epoch 36/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0054 - mae: 0.0457 - val_loss: 102.9815 - val_mae: 6.9201\n", - "Epoch 37/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 1ms/step - loss: 0.0240 - mae: 0.0716 - val_loss: 101.6807 - val_mae: 6.8714\n", - "Epoch 38/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0095 - mae: 0.0490 - val_loss: 101.8840 - val_mae: 6.8803\n", - "Epoch 39/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0027 - mae: 0.0340 - val_loss: 104.6616 - val_mae: 6.9757\n", - "Epoch 40/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0046 - mae: 0.0449 - val_loss: 100.2121 - val_mae: 6.8230\n", - "Epoch 41/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0119 - mae: 0.0650 - val_loss: 102.2873 - val_mae: 6.8914\n", - "Epoch 42/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0276 - mae: 0.0610 - val_loss: 97.7777 - val_mae: 6.6916\n", - "Epoch 43/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0037 - mae: 0.0342 - val_loss: 97.4582 - val_mae: 6.7022\n", - "Epoch 44/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0018 - mae: 0.0285 - val_loss: 97.3804 - val_mae: 6.6951\n", - "Epoch 45/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0029 - mae: 0.0318 - val_loss: 107.3653 - val_mae: 7.2163\n", - "Epoch 46/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0120 - mae: 0.0544 - val_loss: 113.5053 - val_mae: 7.4932\n", - "Epoch 47/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0318 - mae: 0.0865 - val_loss: 92.5976 - val_mae: 6.5065\n", - "Epoch 48/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0011 - mae: 0.0231 - val_loss: 92.7710 - val_mae: 6.5161\n", - "Epoch 49/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0024 - mae: 0.0302 - val_loss: 92.3210 - val_mae: 6.4946\n", - "Epoch 50/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0017 - mae: 0.0271 - val_loss: 91.2108 - val_mae: 6.4525\n", - "Epoch 51/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0079 - mae: 0.0546 - val_loss: 87.0039 - val_mae: 6.2860\n", - "Epoch 52/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0065 - mae: 0.0544 - val_loss: 89.5566 - val_mae: 6.3831\n", - "Epoch 53/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0115 - mae: 0.0557 - val_loss: 88.5476 - val_mae: 6.3611\n", - "Epoch 54/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0386 - mae: 0.1089 - val_loss: 88.2043 - val_mae: 6.3365\n", - "Epoch 55/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0064 - mae: 0.0425 - val_loss: 88.9793 - val_mae: 6.3843\n", - "Epoch 56/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0139 - mae: 0.0683 - val_loss: 86.8086 - val_mae: 6.2762\n", - "Epoch 57/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0621 - mae: 0.0919 - val_loss: 87.9722 - val_mae: 6.3432\n", - "Epoch 58/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0439 - mae: 0.0868 - val_loss: 84.4071 - val_mae: 6.1775\n", - "Epoch 59/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0023 - mae: 0.0279 - val_loss: 85.5060 - val_mae: 6.2414\n", - "Epoch 60/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 0.0021 - mae: 0.0289 - val_loss: 82.9311 - val_mae: 6.1131\n", - "Epoch 61/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0194 - mae: 0.0736 - val_loss: 83.3259 - val_mae: 6.1389\n", - "Epoch 62/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0011 - mae: 0.0237 - val_loss: 83.0009 - val_mae: 6.1206\n", - "Epoch 63/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0060 - mae: 0.0549 - val_loss: 82.8297 - val_mae: 6.1150\n", - "Epoch 64/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0049 - mae: 0.0430 - val_loss: 83.0444 - val_mae: 6.1269\n", - "Epoch 65/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0021 - mae: 0.0314 - val_loss: 80.2624 - val_mae: 6.0063\n", - "Epoch 66/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0138 - mae: 0.0621 - val_loss: 79.2061 - val_mae: 5.9734\n", - "Epoch 67/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0668 - mae: 0.1385 - val_loss: 76.1483 - val_mae: 5.8412\n", - "Epoch 68/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0044 - mae: 0.0339 - val_loss: 76.0317 - val_mae: 5.8318\n", - "Epoch 69/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0368 - mae: 0.0953 - val_loss: 74.2641 - val_mae: 5.7536\n", - "Epoch 70/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0012 - mae: 0.0236 - val_loss: 75.4134 - val_mae: 5.8092\n", - "Epoch 71/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0027 - mae: 0.0311 - val_loss: 76.4617 - val_mae: 5.8594\n", - "Epoch 72/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0012 - mae: 0.0244 - val_loss: 75.3714 - val_mae: 5.8095\n", - "Epoch 73/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0018 - mae: 0.0269 - val_loss: 75.0531 - val_mae: 5.8171\n", - "Epoch 74/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0241 - mae: 0.0899 - val_loss: 72.9266 - val_mae: 5.6904\n", - "Epoch 75/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0021 - mae: 0.0284 - val_loss: 74.0345 - val_mae: 5.7527\n", - "Epoch 76/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0065 - mae: 0.0478 - val_loss: 72.7802 - val_mae: 5.6944\n", - "Epoch 77/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0052 - mae: 0.0444 - val_loss: 77.8162 - val_mae: 5.9745\n", - "Epoch 78/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0033 - mae: 0.0353 - val_loss: 75.6085 - val_mae: 5.8325\n", - "Epoch 79/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0073 - mae: 0.0497 - val_loss: 73.8412 - val_mae: 5.7646\n", - "Epoch 80/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0193 - mae: 0.0812 - val_loss: 73.9291 - val_mae: 5.7575\n", - "Epoch 81/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 1ms/step - loss: 0.0035 - mae: 0.0396 - val_loss: 70.7415 - val_mae: 5.6083\n", - "Epoch 82/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0098 - mae: 0.0630 - val_loss: 71.7451 - val_mae: 5.6640\n", - "Epoch 83/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.1104 - mae: 0.1069 - val_loss: 67.2451 - val_mae: 5.4483\n", - "Epoch 84/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0043 - mae: 0.0328 - val_loss: 66.6703 - val_mae: 5.4243\n", - "Epoch 85/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0038 - mae: 0.0339 - val_loss: 67.4976 - val_mae: 5.4655\n", - "Epoch 86/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 1ms/step - loss: 0.0017 - mae: 0.0301 - val_loss: 67.3472 - val_mae: 5.4626\n", - "Epoch 87/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0014 - mae: 0.0241 - val_loss: 68.7216 - val_mae: 5.5407\n", - "Epoch 88/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0059 - mae: 0.0443 - val_loss: 68.3926 - val_mae: 5.5222\n", - "Epoch 89/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0052 - mae: 0.0418 - val_loss: 65.4597 - val_mae: 5.3872\n", - "Epoch 90/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0268 - mae: 0.0864 - val_loss: 63.5677 - val_mae: 5.2805\n", - "Epoch 91/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0054 - mae: 0.0449 - val_loss: 67.4609 - val_mae: 5.4889\n", - "Epoch 92/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0022 - mae: 0.0328 - val_loss: 64.8628 - val_mae: 5.3513\n", - "Epoch 93/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0417 - mae: 0.0567 - val_loss: 59.6738 - val_mae: 5.1650\n", - "Epoch 94/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0306 - mae: 0.0852 - val_loss: 59.6874 - val_mae: 5.1297\n", - "Epoch 95/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 9.3728e-04 - mae: 0.0196 - val_loss: 60.0518 - val_mae: 5.1552\n", - "Epoch 96/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0013 - mae: 0.0251 - val_loss: 59.8709 - val_mae: 5.1417\n", - "Epoch 97/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0012 - mae: 0.0236 - val_loss: 59.7623 - val_mae: 5.1366\n", - "Epoch 98/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0014 - mae: 0.0254 - val_loss: 59.8039 - val_mae: 5.1426\n", - "Epoch 99/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0012 - mae: 0.0238 - val_loss: 60.0678 - val_mae: 5.1553\n", - "Epoch 100/100\n", - "\u001b[1m370/370\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 1ms/step - loss: 0.0025 - mae: 0.0360 - val_loss: 57.4204 - val_mae: 5.0118\n", - "\u001b[1m548/548\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 1ms/step\n", - "Neural Network MSE: 200.87\n", - "Neural Network MAE: 9.32\n", - " Actual Predicted\n", - "0 72.21 81.753296\n", - "1 71.86 82.512230\n", - "2 71.94 80.226189\n", - "3 71.77 79.577332\n", - "4 71.73 79.338356\n" - ] - } - ], - "source": [ - "import pandas as pd\n", - "import numpy as np\n", - "from sklearn.preprocessing import StandardScaler\n", - "from sklearn.metrics import mean_squared_error, mean_absolute_error\n", - "from tensorflow.keras.models import Sequential\n", - "from tensorflow.keras.layers import Dense\n", - "from tensorflow.keras.callbacks import EarlyStopping\n", - "\n", - "# Load the training and testing data\n", - "training_data = pd.read_csv(\"3_month_training_data.csv\")\n", - "testing_data = pd.read_csv(\"3_month_testing_data.csv\")\n", - "\n", - "# Preprocess data\n", - "training_data = training_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "testing_data = testing_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "\n", - "# Separate features and target\n", - "X_train = training_data.drop(columns=[\"Close\"]).values\n", - "y_train = training_data[\"Close\"].values\n", - "X_test = testing_data.drop(columns=[\"Close\"]).values\n", - "y_test = testing_data[\"Close\"].values\n", - "\n", - "# Standardize the features\n", - "scaler = StandardScaler()\n", - "X_train = scaler.fit_transform(X_train)\n", - "X_test = scaler.transform(X_test)\n", - "\n", - "# Build the neural network model\n", - "model = Sequential([\n", - " Dense(64, activation='relu', input_shape=(X_train.shape[1],)),\n", - " Dense(32, activation='relu'),\n", - " Dense(16, activation='relu'),\n", - " Dense(1) # Output layer for regression\n", - "])\n", - "\n", - "# Compile the model\n", - "model.compile(optimizer='adam', loss='mse', metrics=['mae'])\n", - "\n", - "# Use early stopping to prevent overfitting\n", - "early_stopping = EarlyStopping(monitor='val_loss', patience=10, restore_best_weights=True)\n", - "\n", - "# Train the model\n", - "history = model.fit(\n", - " X_train, y_train,\n", - " epochs=100, # Increase epochs if necessary\n", - " batch_size=32,\n", - " validation_split=0.2, # Use 20% of training data for validation\n", - " callbacks=[early_stopping],\n", - " verbose=1\n", - ")\n", - "\n", - "# Evaluate the model on the test set\n", - "y_pred = model.predict(X_test).flatten()\n", - "mse = mean_squared_error(y_test, y_pred)\n", - "mae = mean_absolute_error(y_test, y_pred)\n", - "\n", - "print(f\"Neural Network MSE: {mse:.2f}\")\n", - "print(f\"Neural Network MAE: {mae:.2f}\")\n", - "\n", - "# Optionally, view a few predictions\n", - "predictions = pd.DataFrame({\"Actual\": y_test, \"Predicted\": y_pred})\n", - "print(predictions.head())" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "id": "89a6efa9-139e-40b3-80ee-a2137d2d1745", - "metadata": {}, - "outputs": [], - "source": [ - "## This neural net is not predicting well.\n", - "## Trying more layers and adjusting parameters to achieve better performace.\n", - "## After I will increase training data set" - ] - }, - { - "cell_type": "code", - "execution_count": 65, - "id": "b3aa2ba5-f2bd-4c01-b699-ace7a25f4f92", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 1/200\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "C:\\Users\\gwitt\\anaconda3\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n", - " super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 3ms/step - loss: 3852.1824 - mae: 61.9999 - val_loss: 3515.2375 - val_mae: 59.2741 - learning_rate: 0.0010\n", - "Epoch 2/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 2753.3269 - mae: 52.4275 - val_loss: 2607.4280 - val_mae: 51.0452 - learning_rate: 0.0010\n", - "Epoch 3/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1982.6472 - mae: 44.4829 - val_loss: 1913.1995 - val_mae: 43.7194 - learning_rate: 0.0010\n", - "Epoch 4/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1394.3098 - mae: 37.2939 - val_loss: 1371.3447 - val_mae: 37.0071 - learning_rate: 0.0010\n", - "Epoch 5/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 947.0174 - mae: 30.7234 - val_loss: 954.2966 - val_mae: 30.8623 - learning_rate: 0.0010\n", - "Epoch 6/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 614.4689 - mae: 24.7321 - val_loss: 641.4285 - val_mae: 25.2906 - learning_rate: 0.0010\n", - "Epoch 7/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 374.7214 - mae: 19.2935 - val_loss: 414.7923 - val_mae: 20.3218 - learning_rate: 0.0010\n", - "Epoch 8/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 212.7006 - mae: 14.5086 - val_loss: 258.0305 - val_mae: 16.0067 - learning_rate: 0.0010\n", - "Epoch 9/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 109.8452 - mae: 10.3872 - val_loss: 155.5748 - val_mae: 12.4000 - learning_rate: 0.0010\n", - "Epoch 10/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 49.6256 - mae: 6.9253 - val_loss: 92.8716 - val_mae: 9.5424 - learning_rate: 0.0010\n", - "Epoch 11/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 19.7037 - mae: 4.2674 - val_loss: 57.3321 - val_mae: 7.4510 - learning_rate: 0.0010\n", - "Epoch 12/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 6.9152 - mae: 2.3767 - val_loss: 38.6574 - val_mae: 6.0698 - learning_rate: 0.0010\n", - "Epoch 13/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 2.5601 - mae: 1.3588 - val_loss: 29.6614 - val_mae: 5.2769 - learning_rate: 0.0010\n", - "Epoch 14/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.4430 - mae: 0.9834 - val_loss: 25.7728 - val_mae: 4.8946 - learning_rate: 0.0010\n", - "Epoch 15/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2715 - mae: 0.8970 - val_loss: 24.3920 - val_mae: 4.7515 - learning_rate: 0.0010\n", - "Epoch 16/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2391 - mae: 0.8785 - val_loss: 24.0039 - val_mae: 4.7105 - learning_rate: 0.0010\n", - "Epoch 17/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2498 - mae: 0.8858 - val_loss: 23.8205 - val_mae: 4.6910 - learning_rate: 0.0010\n", - "Epoch 18/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2596 - mae: 0.8816 - val_loss: 23.7533 - val_mae: 4.6838 - learning_rate: 0.0010\n", - "Epoch 19/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2455 - mae: 0.8761 - val_loss: 23.9365 - val_mae: 4.7033 - learning_rate: 0.0010\n", - "Epoch 20/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2514 - mae: 0.8766 - val_loss: 23.9719 - val_mae: 4.7071 - learning_rate: 0.0010\n", - "Epoch 21/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2481 - mae: 0.8838 - val_loss: 23.8728 - val_mae: 4.6966 - learning_rate: 0.0010\n", - "Epoch 22/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.1993 - mae: 0.8613 - val_loss: 23.9085 - val_mae: 4.7004 - learning_rate: 0.0010\n", - "Epoch 23/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2638 - mae: 0.8847 - val_loss: 23.7717 - val_mae: 4.6858 - learning_rate: 0.0010\n", - "Epoch 24/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2440 - mae: 0.8792 - val_loss: 23.7702 - val_mae: 4.6856 - learning_rate: 5.0000e-04\n", - "Epoch 25/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2284 - mae: 0.8723 - val_loss: 23.7982 - val_mae: 4.6886 - learning_rate: 5.0000e-04\n", - "Epoch 26/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2577 - mae: 0.8822 - val_loss: 23.8829 - val_mae: 4.6976 - learning_rate: 5.0000e-04\n", - "Epoch 27/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2404 - mae: 0.8792 - val_loss: 23.9080 - val_mae: 4.7003 - learning_rate: 5.0000e-04\n", - "Epoch 28/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2250 - mae: 0.8717 - val_loss: 23.7182 - val_mae: 4.6801 - learning_rate: 5.0000e-04\n", - "Epoch 29/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2458 - mae: 0.8778 - val_loss: 23.9344 - val_mae: 4.7031 - learning_rate: 5.0000e-04\n", - "Epoch 30/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2029 - mae: 0.8612 - val_loss: 23.9050 - val_mae: 4.7000 - learning_rate: 5.0000e-04\n", - "Epoch 31/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2488 - mae: 0.8808 - val_loss: 24.2789 - val_mae: 4.7396 - learning_rate: 5.0000e-04\n", - "Epoch 32/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2553 - mae: 0.8822 - val_loss: 23.6859 - val_mae: 4.6766 - learning_rate: 5.0000e-04\n", - "Epoch 33/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 3ms/step - loss: 1.2188 - mae: 0.8690 - val_loss: 24.0428 - val_mae: 4.7146 - learning_rate: 5.0000e-04\n", - "Epoch 34/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2323 - mae: 0.8778 - val_loss: 23.4817 - val_mae: 4.6547 - learning_rate: 5.0000e-04\n", - "Epoch 35/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2333 - mae: 0.8731 - val_loss: 23.9849 - val_mae: 4.7085 - learning_rate: 5.0000e-04\n", - "Epoch 36/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2414 - mae: 0.8759 - val_loss: 23.6787 - val_mae: 4.6759 - learning_rate: 5.0000e-04\n", - "Epoch 37/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2447 - mae: 0.8826 - val_loss: 23.5780 - val_mae: 4.6651 - learning_rate: 5.0000e-04\n", - "Epoch 38/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2364 - mae: 0.8758 - val_loss: 23.8652 - val_mae: 4.6958 - learning_rate: 5.0000e-04\n", - "Epoch 39/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 1.2106 - mae: 0.8608 - val_loss: 16.9429 - val_mae: 3.8901 - learning_rate: 5.0000e-04\n", - "Epoch 40/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.2149 - mae: 0.3297 - val_loss: 12.0071 - val_mae: 3.1946 - learning_rate: 5.0000e-04\n", - "Epoch 41/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.1148 - mae: 0.2440 - val_loss: 10.3470 - val_mae: 2.9257 - learning_rate: 5.0000e-04\n", - "Epoch 42/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0872 - mae: 0.2155 - val_loss: 9.2182 - val_mae: 2.7275 - learning_rate: 5.0000e-04\n", - "Epoch 43/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0760 - mae: 0.2020 - val_loss: 8.3651 - val_mae: 2.5716 - learning_rate: 5.0000e-04\n", - "Epoch 44/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0650 - mae: 0.1867 - val_loss: 7.5653 - val_mae: 2.4151 - learning_rate: 5.0000e-04\n", - "Epoch 45/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0548 - mae: 0.1731 - val_loss: 6.8565 - val_mae: 2.2615 - learning_rate: 5.0000e-04\n", - "Epoch 46/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0484 - mae: 0.1628 - val_loss: 6.0878 - val_mae: 2.0981 - learning_rate: 5.0000e-04\n", - "Epoch 47/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0410 - mae: 0.1540 - val_loss: 5.4892 - val_mae: 1.9504 - learning_rate: 5.0000e-04\n", - "Epoch 48/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0372 - mae: 0.1460 - val_loss: 5.3494 - val_mae: 1.9196 - learning_rate: 5.0000e-04\n", - "Epoch 49/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0338 - mae: 0.1408 - val_loss: 5.1110 - val_mae: 1.8529 - learning_rate: 5.0000e-04\n", - "Epoch 50/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0313 - mae: 0.1333 - val_loss: 4.9872 - val_mae: 1.8187 - learning_rate: 5.0000e-04\n", - "Epoch 51/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0299 - mae: 0.1311 - val_loss: 4.9263 - val_mae: 1.8014 - learning_rate: 5.0000e-04\n", - "Epoch 52/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0290 - mae: 0.1277 - val_loss: 4.9590 - val_mae: 1.8068 - learning_rate: 5.0000e-04\n", - "Epoch 53/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0271 - mae: 0.1231 - val_loss: 4.8990 - val_mae: 1.7905 - learning_rate: 5.0000e-04\n", - "Epoch 54/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0253 - mae: 0.1196 - val_loss: 5.0170 - val_mae: 1.8210 - learning_rate: 5.0000e-04\n", - "Epoch 55/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0238 - mae: 0.1175 - val_loss: 4.9177 - val_mae: 1.7945 - learning_rate: 5.0000e-04\n", - "Epoch 56/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0220 - mae: 0.1124 - val_loss: 4.9086 - val_mae: 1.7923 - learning_rate: 5.0000e-04\n", - "Epoch 57/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0208 - mae: 0.1110 - val_loss: 4.9697 - val_mae: 1.8064 - learning_rate: 5.0000e-04\n", - "Epoch 58/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0212 - mae: 0.1109 - val_loss: 5.0626 - val_mae: 1.8284 - learning_rate: 5.0000e-04\n", - "Epoch 59/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0215 - mae: 0.1072 - val_loss: 5.0135 - val_mae: 1.8166 - learning_rate: 2.5000e-04\n", - "Epoch 60/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0189 - mae: 0.1056 - val_loss: 4.9738 - val_mae: 1.8071 - learning_rate: 2.5000e-04\n", - "Epoch 61/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0200 - mae: 0.1063 - val_loss: 4.9922 - val_mae: 1.8114 - learning_rate: 2.5000e-04\n", - "Epoch 62/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0183 - mae: 0.1045 - val_loss: 4.9533 - val_mae: 1.8020 - learning_rate: 2.5000e-04\n", - "Epoch 63/200\n", - "\u001b[1m423/423\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step - loss: 0.0180 - mae: 0.1016 - val_loss: 5.0181 - val_mae: 1.8176 - learning_rate: 2.5000e-04\n", - "\u001b[1m548/548\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step\n", - "Neural Network MSE: 7.02\n", - "Neural Network MAE: 1.84\n" - ] - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "import pandas as pd\n", - "import numpy as np\n", - "from sklearn.preprocessing import StandardScaler\n", - "from sklearn.metrics import mean_squared_error, mean_absolute_error\n", - "from tensorflow.keras.models import Sequential\n", - "from tensorflow.keras.layers import Dense, Dropout\n", - "from tensorflow.keras.callbacks import EarlyStopping, ReduceLROnPlateau\n", - "import matplotlib.pyplot as plt\n", - "\n", - "# Load the training and testing data\n", - "training_data = pd.read_csv(\"3_month_training_data.csv\")\n", - "testing_data = pd.read_csv(\"3_month_testing_data.csv\")\n", - "\n", - "# Preprocess data\n", - "training_data = training_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "testing_data = testing_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "\n", - "# Separate features and target\n", - "X_train = training_data.drop(columns=[\"Close\"]).values\n", - "y_train = training_data[\"Close\"].values\n", - "X_test = testing_data.drop(columns=[\"Close\"]).values\n", - "y_test = testing_data[\"Close\"].values\n", - "\n", - "# Standardize the features\n", - "scaler = StandardScaler()\n", - "X_train = scaler.fit_transform(X_train)\n", - "X_test = scaler.transform(X_test)\n", - "\n", - "# Build the neural network model with dropout and reduced learning rate\n", - "model = Sequential([\n", - " Dense(128, activation='sigmoid', input_shape=(X_train.shape[1],)),\n", - " Dropout(0.2), # Dropout layer for regularization\n", - " Dense(64, activation='sigmoid'),\n", - " Dropout(0.2),\n", - " Dense(32, activation='sigmoid'),\n", - " Dense(1) # Output layer for regression\n", - "])\n", - "\n", - "# Compile the model with a reduced learning rate\n", - "model.compile(optimizer='adam', loss='mse', metrics=['mae'])\n", - "\n", - "# Callbacks for early stopping and reducing learning rate on plateau\n", - "early_stopping = EarlyStopping(monitor='val_loss', patience=10, restore_best_weights=True)\n", - "reduce_lr = ReduceLROnPlateau(monitor='val_loss', factor=0.5, patience=5, min_lr=1e-6)\n", - "\n", - "# Train the model with callbacks\n", - "history = model.fit(\n", - " X_train, y_train,\n", - " epochs=200, # Increase epochs for better convergence\n", - " batch_size=32,\n", - " validation_split=0.2,\n", - " callbacks=[early_stopping, reduce_lr],\n", - " verbose=1\n", - ")\n", - "\n", - "# Evaluate the model on the test set\n", - "y_pred = model.predict(X_test).flatten()\n", - "mse = mean_squared_error(y_test, y_pred)\n", - "mae = mean_absolute_error(y_test, y_pred)\n", - "\n", - "print(f\"Neural Network MSE: {mse:.2f}\")\n", - "print(f\"Neural Network MAE: {mae:.2f}\")\n", - "\n", - "# Visualization for Actual vs Predicted\n", - "plt.figure(figsize=(12, 6))\n", - "plt.plot(y_test, label=\"Actual\", color=\"blue\")\n", - "plt.plot(y_pred, label=\"Predicted\", color=\"red\", linestyle=\"dashed\")\n", - "plt.title(\"Actual vs Predicted Close Prices\")\n", - "plt.xlabel(\"Time Steps\")\n", - "plt.ylabel(\"Close Price\")\n", - "plt.legend()\n", - "plt.show()\n", - "\n", - "# Visualization of Training and Validation Loss\n", - "plt.figure(figsize=(10, 5))\n", - "plt.plot(history.history['loss'], label='Training Loss')\n", - "plt.plot(history.history['val_loss'], label='Validation Loss')\n", - "plt.title(\"Training and Validation Loss Over Epochs\")\n", - "plt.xlabel(\"Epochs\")\n", - "plt.ylabel(\"Loss (MSE)\")\n", - "plt.legend()\n", - "plt.show()" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "id": "7f28be3f-d894-4987-983c-10d1df5b336a", - "metadata": {}, - "outputs": [], - "source": [ - "## Now we are going to extract a larger training data set." - ] - }, - { - "cell_type": "code", - "execution_count": 73, - "id": "f3ab7c8f-57ab-476a-bc9e-12057a86e237", - "metadata": {}, - "outputs": [ - { - "name": "stderr", - "output_type": "stream", - "text": [ - "ERROR -1 2104 Market data farm connection is OK:usfarm.nj\n", - "ERROR -1 2104 Market data farm connection is OK:usfuture\n", - "ERROR -1 2104 Market data farm connection is OK:cashfarm\n", - "ERROR -1 2104 Market data farm connection is OK:usfarm\n", - "ERROR -1 2106 HMDS data farm connection is OK:ushmds\n", - "ERROR -1 2158 Sec-def data farm connection is OK:secdefnj\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Historical Data Ended\n", - "Time taken to pull data: 678.25 seconds\n", - " Date Open High Low Close Volume\n", - "0 20200803 10:50:00 47.13 47.13 47.13 47.13 1\n", - "1 20200803 10:55:00 47.13 47.13 47.13 47.13 4\n", - "2 20200803 11:00:00 47.18 47.18 47.18 47.18 1\n", - "3 20200803 11:05:00 47.18 47.18 47.18 47.18 8\n", - "4 20200803 11:10:00 47.18 47.18 47.18 47.18 0\n" - ] - } - ], - "source": [ - "## TRAINING DATA PULL (July 2021 - July 2024 - 5 mins - 3 years of data)\n", - "\n", - "from ibapi.client import EClient\n", - "from ibapi.wrapper import EWrapper\n", - "from ibapi.contract import Contract\n", - "import threading\n", - "import time\n", - "import pandas as pd\n", - "\n", - "class IBApi(EWrapper, EClient):\n", - " def __init__(self):\n", - " EClient.__init__(self, self)\n", - " self.data = [] # Store data\n", - " self.data_retrieved = False # Flag to check if data retrieval is complete\n", - "\n", - " def historicalData(self, reqId, bar):\n", - " self.data.append({\n", - " \"Date\": bar.date,\n", - " \"Open\": bar.open,\n", - " \"High\": bar.high,\n", - " \"Low\": bar.low,\n", - " \"Close\": bar.close,\n", - " \"Volume\": bar.volume\n", - " })\n", - "\n", - " def historicalDataEnd(self, reqId, start, end):\n", - " print(\"Historical Data Ended\")\n", - " self.df = pd.DataFrame(self.data)\n", - " self.data_retrieved = True # Set the flag to True to indicate data retrieval completion\n", - " self.disconnect()\n", - "\n", - "class IBApp:\n", - " def __init__(self):\n", - " self.app = IBApi()\n", - "\n", - " def connect(self):\n", - " self.app.connect(\"127.0.0.1\", 7496, 0)\n", - " thread = threading.Thread(target=self.run_app, daemon=True)\n", - " thread.start()\n", - " time.sleep(1)\n", - "\n", - " def run_app(self):\n", - " self.app.run()\n", - "\n", - " def request_training_data(self):\n", - " contract = Contract()\n", - " contract.symbol = \"CL\"\n", - " contract.secType = \"FUT\"\n", - " contract.exchange = \"NYMEX\"\n", - " contract.currency = \"USD\"\n", - " contract.lastTradeDateOrContractMonth = \"202412\" # November 2024 contract\n", - "\n", - " # Set parameters for data pull\n", - " end_date = \"20230730 23:59:59 UTC\" # Example end date in UTC\n", - " duration = \"3 Y\" # 5 Year Duration\n", - " bar_size = \"5 mins\"\n", - "\n", - " # Record start time\n", - " start_time = time.time()\n", - "\n", - " # Request historical data\n", - " self.app.reqHistoricalData(\n", - " reqId=1,\n", - " contract=contract,\n", - " endDateTime=end_date,\n", - " durationStr=duration,\n", - " barSizeSetting=bar_size,\n", - " whatToShow='TRADES',\n", - " useRTH=0,\n", - " formatDate=1,\n", - " keepUpToDate=False,\n", - " chartOptions=[]\n", - " )\n", - "\n", - " # Wait until data retrieval is complete\n", - " while not self.app.data_retrieved:\n", - " time.sleep(0.1) # Small sleep interval to prevent busy-waiting\n", - "\n", - " # Record end time and calculate elapsed time\n", - " end_time = time.time()\n", - " elapsed_time = end_time - start_time\n", - " print(f\"Time taken to pull data: {elapsed_time:.2f} seconds\")\n", - "\n", - " def disconnect(self):\n", - " self.app.disconnect()\n", - "\n", - "# Instantiate and connect the app\n", - "app = IBApp()\n", - "app.connect()\n", - "\n", - "# Request training data\n", - "app.request_training_data()\n", - "\n", - "# Access the DataFrame\n", - "train_data = app.app.df if hasattr(app.app, 'df') else pd.DataFrame()\n", - "\n", - "# Disconnect from API\n", - "app.disconnect()\n", - "\n", - "# Display the training data\n", - "print(train_data.head())" - ] - }, - { - "cell_type": "code", - "execution_count": 78, - "id": "00bd8704-8ead-42c2-b270-fd08cf8cff04", - "metadata": {}, - "outputs": [], - "source": [ - "train_data.to_csv(\"3_years_training_data.csv\")" - ] - }, - { - "cell_type": "code", - "execution_count": null, - "id": "281d3e0d-d3b5-4055-9a6e-748392330c39", - "metadata": {}, - "outputs": [], - "source": [ - "# now lets run the same tests using a much larger training set" - ] - }, - { - "cell_type": "code", - "execution_count": 84, - "id": "756d2478-f21f-46bb-bce1-64b00183e2a1", - "metadata": { - "scrolled": true - }, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 1/100\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "C:\\Users\\gwitt\\anaconda3\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n", - " super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 2ms/step - loss: 572.8583 - mae: 11.6309 - val_loss: 0.3461 - val_mae: 0.1569\n", - "Epoch 2/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 0.0180 - mae: 0.0547 - val_loss: 0.0547 - val_mae: 0.0683\n", - "Epoch 3/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m9s\u001b[0m 3ms/step - loss: 0.0236 - mae: 0.0682 - val_loss: 3.1116 - val_mae: 0.4845\n", - "Epoch 4/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0340 - mae: 0.0836 - val_loss: 0.0266 - val_mae: 0.0549\n", - "Epoch 5/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 0.0390 - mae: 0.0687 - val_loss: 0.2596 - val_mae: 0.1420\n", - "Epoch 6/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0235 - mae: 0.0592 - val_loss: 0.0114 - val_mae: 0.0722\n", - "Epoch 7/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0197 - mae: 0.0535 - val_loss: 0.0157 - val_mae: 0.0416\n", - "Epoch 8/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0125 - mae: 0.0491 - val_loss: 0.0114 - val_mae: 0.0532\n", - "Epoch 9/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0170 - mae: 0.0565 - val_loss: 0.0331 - val_mae: 0.1077\n", - "Epoch 10/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0219 - mae: 0.0597 - val_loss: 0.0243 - val_mae: 0.1437\n", - "Epoch 11/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 0.0309 - mae: 0.0763 - val_loss: 0.0086 - val_mae: 0.0536\n", - "Epoch 12/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0138 - mae: 0.0563 - val_loss: 0.0055 - val_mae: 0.0429\n", - "Epoch 13/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0341 - mae: 0.0587 - val_loss: 0.0054 - val_mae: 0.0392\n", - "Epoch 14/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0176 - mae: 0.0603 - val_loss: 0.0154 - val_mae: 0.0908\n", - "Epoch 15/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0157 - mae: 0.0450 - val_loss: 0.0139 - val_mae: 0.0770\n", - "Epoch 16/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0090 - mae: 0.0389 - val_loss: 0.0024 - val_mae: 0.0353\n", - "Epoch 17/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0404 - mae: 0.0745 - val_loss: 0.0016 - val_mae: 0.0223\n", - "Epoch 18/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0112 - mae: 0.0534 - val_loss: 0.0025 - val_mae: 0.0219\n", - "Epoch 19/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0183 - mae: 0.0470 - val_loss: 0.0137 - val_mae: 0.1113\n", - "Epoch 20/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0070 - mae: 0.0544 - val_loss: 0.0050 - val_mae: 0.0377\n", - "Epoch 21/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0153 - mae: 0.0547 - val_loss: 0.0049 - val_mae: 0.0296\n", - "Epoch 22/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0102 - mae: 0.0359 - val_loss: 0.0393 - val_mae: 0.0476\n", - "Epoch 23/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0027 - mae: 0.0267 - val_loss: 0.0416 - val_mae: 0.0737\n", - "Epoch 24/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0260 - mae: 0.0696 - val_loss: 0.0192 - val_mae: 0.0371\n", - "Epoch 25/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 0.0099 - mae: 0.0366 - val_loss: 0.0047 - val_mae: 0.0575\n", - "Epoch 26/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0071 - mae: 0.0430 - val_loss: 0.0263 - val_mae: 0.1461\n", - "Epoch 27/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0072 - mae: 0.0440 - val_loss: 0.0172 - val_mae: 0.0845\n", - "\u001b[1m548/548\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 2ms/step\n", - "Neural Network MSE: 7.84\n", - "Neural Network MAE: 0.24\n", - " Actual Predicted\n", - "0 72.21 72.106239\n", - "1 71.86 71.874084\n", - "2 71.94 71.881981\n", - "3 71.77 71.743423\n", - "4 71.73 71.717209\n" - ] - } - ], - "source": [ - "# relu method\n", - "\n", - "import pandas as pd\n", - "import numpy as np\n", - "from sklearn.preprocessing import StandardScaler\n", - "from sklearn.metrics import mean_squared_error, mean_absolute_error\n", - "from tensorflow.keras.models import Sequential\n", - "from tensorflow.keras.layers import Dense\n", - "from tensorflow.keras.callbacks import EarlyStopping\n", - "\n", - "# Load the training and testing data\n", - "training_data = pd.read_csv(\"3_years_training_data.csv\")\n", - "testing_data = pd.read_csv(\"3_month_testing_data.csv\")\n", - "\n", - "# Preprocess data\n", - "training_data = training_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "testing_data = testing_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "\n", - "# Separate features and target\n", - "X_train = training_data.drop(columns=[\"Close\"]).values\n", - "y_train = training_data[\"Close\"].values\n", - "X_test = testing_data.drop(columns=[\"Close\"]).values\n", - "y_test = testing_data[\"Close\"].values\n", - "\n", - "# Standardize the features\n", - "scaler = StandardScaler()\n", - "X_train = scaler.fit_transform(X_train)\n", - "X_test = scaler.transform(X_test)\n", - "\n", - "# Build the neural network model\n", - "model = Sequential([\n", - " Dense(64, activation='relu', input_shape=(X_train.shape[1],)),\n", - " Dense(32, activation='relu'),\n", - " Dense(16, activation='relu'),\n", - " Dense(1) # Output layer for regression\n", - "])\n", - "\n", - "# Compile the model\n", - "model.compile(optimizer='adam', loss='mse', metrics=['mae'])\n", - "\n", - "# Use early stopping to prevent overfitting\n", - "early_stopping = EarlyStopping(monitor='val_loss', patience=10, restore_best_weights=True)\n", - "\n", - "# Train the model\n", - "history = model.fit(\n", - " X_train, y_train,\n", - " epochs=100, # Increase epochs if necessary\n", - " batch_size=32,\n", - " validation_split=0.2, # Use 20% of training data for validation\n", - " callbacks=[early_stopping],\n", - " verbose=1\n", - ")\n", - "\n", - "# Evaluate the model on the test set\n", - "y_pred = model.predict(X_test).flatten()\n", - "mse = mean_squared_error(y_test, y_pred)\n", - "mae = mean_absolute_error(y_test, y_pred)\n", - "\n", - "print(f\"Neural Network MSE: {mse:.2f}\")\n", - "print(f\"Neural Network MAE: {mae:.2f}\")\n", - "\n", - "# Optionally, view a few predictions\n", - "predictions = pd.DataFrame({\"Actual\": y_test, \"Predicted\": y_pred})\n", - "print(predictions.head())" - ] - }, - { - "cell_type": "code", - "execution_count": 86, - "id": "e12364e0-b3ee-4deb-a6e0-307e4affa44d", - "metadata": {}, - "outputs": [ - { - "name": "stdout", - "output_type": "stream", - "text": [ - "Epoch 1/100\n" - ] - }, - { - "name": "stderr", - "output_type": "stream", - "text": [ - "C:\\Users\\gwitt\\anaconda3\\Lib\\site-packages\\keras\\src\\layers\\core\\dense.py:87: UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. When using Sequential models, prefer using an `Input(shape)` object as the first layer in the model instead.\n", - " super().__init__(activity_regularizer=activity_regularizer, **kwargs)\n" - ] - }, - { - "name": "stdout", - "output_type": "stream", - "text": [ - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 2ms/step - loss: 3297.1851 - mae: 56.0110 - val_loss: 1020.7242 - val_mae: 31.8577\n", - "Epoch 2/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 693.4044 - mae: 24.0952 - val_loss: 62.8939 - val_mae: 7.5552\n", - "Epoch 3/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 83.8230 - mae: 7.4161 - val_loss: 0.0635 - val_mae: 0.1529\n", - "Epoch 4/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 1.6463 - mae: 0.4935 - val_loss: 0.0113 - val_mae: 0.0622\n", - "Epoch 5/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 0.0907 - mae: 0.0778 - val_loss: 0.0125 - val_mae: 0.0966\n", - "Epoch 6/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 0.0057 - mae: 0.0355 - val_loss: 0.0071 - val_mae: 0.0654\n", - "Epoch 7/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0020 - mae: 0.0287 - val_loss: 0.0049 - val_mae: 0.0308\n", - "Epoch 8/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 0.0016 - mae: 0.0262 - val_loss: 0.0044 - val_mae: 0.0338\n", - "Epoch 9/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0016 - mae: 0.0260 - val_loss: 0.0021 - val_mae: 0.0202\n", - "Epoch 10/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 0.0015 - mae: 0.0251 - val_loss: 0.0020 - val_mae: 0.0221\n", - "Epoch 11/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 0.0015 - mae: 0.0239 - val_loss: 0.0036 - val_mae: 0.0332\n", - "Epoch 12/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 0.0015 - mae: 0.0250 - val_loss: 0.0018 - val_mae: 0.0232\n", - "Epoch 13/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0014 - mae: 0.0237 - val_loss: 0.0020 - val_mae: 0.0211\n", - "Epoch 14/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0014 - mae: 0.0246 - val_loss: 0.0018 - val_mae: 0.0215\n", - "Epoch 15/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0013 - mae: 0.0231 - val_loss: 0.0018 - val_mae: 0.0260\n", - "Epoch 16/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0013 - mae: 0.0230 - val_loss: 0.0026 - val_mae: 0.0348\n", - "Epoch 17/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0014 - mae: 0.0236 - val_loss: 0.0021 - val_mae: 0.0263\n", - "Epoch 18/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 1ms/step - loss: 0.0013 - mae: 0.0232 - val_loss: 0.0051 - val_mae: 0.0280\n", - "Epoch 19/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0014 - mae: 0.0241 - val_loss: 0.0020 - val_mae: 0.0288\n", - "Epoch 20/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0012 - mae: 0.0226 - val_loss: 0.0016 - val_mae: 0.0195\n", - "Epoch 21/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0013 - mae: 0.0225 - val_loss: 0.0017 - val_mae: 0.0250\n", - "Epoch 22/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0013 - mae: 0.0227 - val_loss: 0.0030 - val_mae: 0.0450\n", - "Epoch 23/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0012 - mae: 0.0228 - val_loss: 0.0066 - val_mae: 0.0518\n", - "Epoch 24/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0013 - mae: 0.0227 - val_loss: 0.0029 - val_mae: 0.0294\n", - "Epoch 25/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 0.0011 - mae: 0.0219 - val_loss: 0.0015 - val_mae: 0.0209\n", - "Epoch 26/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m7s\u001b[0m 2ms/step - loss: 0.0011 - mae: 0.0205 - val_loss: 0.0019 - val_mae: 0.0258\n", - "Epoch 27/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 0.0012 - mae: 0.0223 - val_loss: 0.0018 - val_mae: 0.0275\n", - "Epoch 28/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0011 - mae: 0.0212 - val_loss: 0.0014 - val_mae: 0.0226\n", - "Epoch 29/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 0.0011 - mae: 0.0213 - val_loss: 0.0015 - val_mae: 0.0256\n", - "Epoch 30/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0011 - mae: 0.0214 - val_loss: 0.0018 - val_mae: 0.0292\n", - "Epoch 31/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0011 - mae: 0.0209 - val_loss: 0.0032 - val_mae: 0.0346\n", - "Epoch 32/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0011 - mae: 0.0211 - val_loss: 0.0013 - val_mae: 0.0190\n", - "Epoch 33/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0010 - mae: 0.0213 - val_loss: 0.0021 - val_mae: 0.0343\n", - "Epoch 34/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 9.9637e-04 - mae: 0.0203 - val_loss: 0.0027 - val_mae: 0.0423\n", - "Epoch 35/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0010 - mae: 0.0208 - val_loss: 0.0021 - val_mae: 0.0289\n", - "Epoch 36/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 9.3876e-04 - mae: 0.0194 - val_loss: 0.0013 - val_mae: 0.0218\n", - "Epoch 37/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0010 - mae: 0.0203 - val_loss: 0.0024 - val_mae: 0.0387\n", - "Epoch 38/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 9.6806e-04 - mae: 0.0205 - val_loss: 0.0016 - val_mae: 0.0196\n", - "Epoch 39/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0010 - mae: 0.0203 - val_loss: 0.0016 - val_mae: 0.0217\n", - "Epoch 40/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 9.8292e-04 - mae: 0.0207 - val_loss: 0.0012 - val_mae: 0.0180\n", - "Epoch 41/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 9.7501e-04 - mae: 0.0199 - val_loss: 0.0018 - val_mae: 0.0268\n", - "Epoch 42/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 0.0011 - mae: 0.0209 - val_loss: 0.0012 - val_mae: 0.0168\n", - "Epoch 43/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.9433e-04 - mae: 0.0191 - val_loss: 0.0016 - val_mae: 0.0290\n", - "Epoch 44/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 9.6032e-04 - mae: 0.0198 - val_loss: 0.0025 - val_mae: 0.0332\n", - "Epoch 45/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 0.0010 - mae: 0.0198 - val_loss: 0.0017 - val_mae: 0.0303\n", - "Epoch 46/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.7278e-04 - mae: 0.0192 - val_loss: 0.0016 - val_mae: 0.0252\n", - "Epoch 47/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 9.4546e-04 - mae: 0.0196 - val_loss: 0.0015 - val_mae: 0.0290\n", - "Epoch 48/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 1ms/step - loss: 8.6356e-04 - mae: 0.0185 - val_loss: 0.0053 - val_mae: 0.0667\n", - "Epoch 49/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.7887e-04 - mae: 0.0193 - val_loss: 0.0011 - val_mae: 0.0179\n", - "Epoch 50/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.3700e-04 - mae: 0.0184 - val_loss: 0.0025 - val_mae: 0.0410\n", - "Epoch 51/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 1ms/step - loss: 8.2044e-04 - mae: 0.0186 - val_loss: 0.0015 - val_mae: 0.0221\n", - "Epoch 52/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.5456e-04 - mae: 0.0183 - val_loss: 0.0021 - val_mae: 0.0225\n", - "Epoch 53/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.0283e-04 - mae: 0.0182 - val_loss: 0.0030 - val_mae: 0.0470\n", - "Epoch 54/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.9499e-04 - mae: 0.0187 - val_loss: 0.0019 - val_mae: 0.0341\n", - "Epoch 55/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.1096e-04 - mae: 0.0185 - val_loss: 0.0043 - val_mae: 0.0314\n", - "Epoch 56/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 8.4730e-04 - mae: 0.0186 - val_loss: 0.0013 - val_mae: 0.0238\n", - "Epoch 57/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 8.5474e-04 - mae: 0.0183 - val_loss: 0.0042 - val_mae: 0.0353\n", - "Epoch 58/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.7424e-04 - mae: 0.0173 - val_loss: 0.0011 - val_mae: 0.0202\n", - "Epoch 59/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.1713e-04 - mae: 0.0171 - val_loss: 9.1875e-04 - val_mae: 0.0172\n", - "Epoch 60/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.5990e-04 - mae: 0.0177 - val_loss: 0.0013 - val_mae: 0.0268\n", - "Epoch 61/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.0998e-04 - mae: 0.0168 - val_loss: 0.0019 - val_mae: 0.0279\n", - "Epoch 62/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.0700e-04 - mae: 0.0175 - val_loss: 0.0012 - val_mae: 0.0196\n", - "Epoch 63/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 1ms/step - loss: 9.5286e-04 - mae: 0.0191 - val_loss: 9.3224e-04 - val_mae: 0.0165\n", - "Epoch 64/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 8.1920e-04 - mae: 0.0182 - val_loss: 0.0012 - val_mae: 0.0265\n", - "Epoch 65/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 7.0302e-04 - mae: 0.0169 - val_loss: 9.5052e-04 - val_mae: 0.0144\n", - "Epoch 66/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.9119e-04 - mae: 0.0183 - val_loss: 0.0023 - val_mae: 0.0364\n", - "Epoch 67/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.3073e-04 - mae: 0.0174 - val_loss: 0.0040 - val_mae: 0.0385\n", - "Epoch 68/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 1ms/step - loss: 7.0992e-04 - mae: 0.0168 - val_loss: 0.0057 - val_mae: 0.0332\n", - "Epoch 69/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.1660e-04 - mae: 0.0173 - val_loss: 8.6993e-04 - val_mae: 0.0149\n", - "Epoch 70/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.9050e-04 - mae: 0.0168 - val_loss: 9.1320e-04 - val_mae: 0.0204\n", - "Epoch 71/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 6.4809e-04 - mae: 0.0164 - val_loss: 8.5441e-04 - val_mae: 0.0193\n", - "Epoch 72/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 6.7284e-04 - mae: 0.0166 - val_loss: 8.6238e-04 - val_mae: 0.0167\n", - "Epoch 73/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m10s\u001b[0m 2ms/step - loss: 6.6893e-04 - mae: 0.0162 - val_loss: 0.0058 - val_mae: 0.0296\n", - "Epoch 74/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.0679e-04 - mae: 0.0165 - val_loss: 0.0010 - val_mae: 0.0207\n", - "Epoch 75/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 5.7697e-04 - mae: 0.0157 - val_loss: 0.0011 - val_mae: 0.0192\n", - "Epoch 76/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 6.2833e-04 - mae: 0.0161 - val_loss: 0.0058 - val_mae: 0.0720\n", - "Epoch 77/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 7.1162e-04 - mae: 0.0167 - val_loss: 0.0032 - val_mae: 0.0494\n", - "Epoch 78/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 6.5035e-04 - mae: 0.0165 - val_loss: 7.1719e-04 - val_mae: 0.0148\n", - "Epoch 79/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 7.3297e-04 - mae: 0.0168 - val_loss: 8.8522e-04 - val_mae: 0.0209\n", - "Epoch 80/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 2ms/step - loss: 6.4260e-04 - mae: 0.0166 - val_loss: 9.8242e-04 - val_mae: 0.0229\n", - "Epoch 81/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m6s\u001b[0m 2ms/step - loss: 6.4046e-04 - mae: 0.0168 - val_loss: 7.8706e-04 - val_mae: 0.0166\n", - "Epoch 82/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m8s\u001b[0m 2ms/step - loss: 6.6381e-04 - mae: 0.0156 - val_loss: 0.0017 - val_mae: 0.0339\n", - "Epoch 83/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 6.6207e-04 - mae: 0.0161 - val_loss: 9.0922e-04 - val_mae: 0.0160\n", - "Epoch 84/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 6.4263e-04 - mae: 0.0158 - val_loss: 8.1381e-04 - val_mae: 0.0163\n", - "Epoch 85/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 6.9406e-04 - mae: 0.0160 - val_loss: 0.0014 - val_mae: 0.0242\n", - "Epoch 86/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 5.9793e-04 - mae: 0.0159 - val_loss: 0.0015 - val_mae: 0.0322\n", - "Epoch 87/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 6.1736e-04 - mae: 0.0155 - val_loss: 0.0022 - val_mae: 0.0399\n", - "Epoch 88/100\n", - "\u001b[1m3443/3443\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 1ms/step - loss: 5.6485e-04 - mae: 0.0153 - val_loss: 0.0016 - val_mae: 0.0337\n", - "\u001b[1m548/548\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m1s\u001b[0m 981us/step\n", - "Neural Network MSE: 0.01\n", - "Neural Network MAE: 0.04\n", - " Actual Predicted\n", - "0 72.21 72.200966\n", - "1 71.86 71.966621\n", - "2 71.94 71.904488\n", - "3 71.77 71.728455\n", - "4 71.73 71.720940\n" - ] - } - ], - "source": [ - "# sigmoid\n", - "\n", - "import pandas as pd\n", - "import numpy as np\n", - "from sklearn.preprocessing import StandardScaler\n", - "from sklearn.metrics import mean_squared_error, mean_absolute_error\n", - "from tensorflow.keras.models import Sequential\n", - "from tensorflow.keras.layers import Dense\n", - "from tensorflow.keras.callbacks import EarlyStopping\n", - "\n", - "# Load the training and testing data\n", - "training_data = pd.read_csv(\"3_years_training_data.csv\")\n", - "testing_data = pd.read_csv(\"3_month_testing_data.csv\")\n", - "\n", - "# Preprocess data\n", - "training_data = training_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "testing_data = testing_data.drop(columns=[\"Unnamed: 0\", \"Date\"])\n", - "\n", - "# Separate features and target\n", - "X_train = training_data.drop(columns=[\"Close\"]).values\n", - "y_train = training_data[\"Close\"].values\n", - "X_test = testing_data.drop(columns=[\"Close\"]).values\n", - "y_test = testing_data[\"Close\"].values\n", - "\n", - "# Standardize the features\n", - "scaler = StandardScaler()\n", - "X_train = scaler.fit_transform(X_train)\n", - "X_test = scaler.transform(X_test)\n", - "\n", - "# Build the neural network model\n", - "model = Sequential([\n", - " Dense(64, activation='sigmoid', input_shape=(X_train.shape[1],)),\n", - " Dense(32, activation='sigmoid'),\n", - " Dense(16, activation='sigmoid'),\n", - " Dense(1) # Output layer for regression\n", - "])\n", - "\n", - "# Compile the model\n", - "model.compile(optimizer='adam', loss='mse', metrics=['mae'])\n", - "\n", - "# Use early stopping to prevent overfitting\n", - "early_stopping = EarlyStopping(monitor='val_loss', patience=10, restore_best_weights=True)\n", - "\n", - "# Train the model\n", - "history = model.fit(\n", - " X_train, y_train,\n", - " epochs=100, # Increase epochs if necessary\n", - " batch_size=32,\n", - " validation_split=0.2, # Use 20% of training data for validation\n", - " callbacks=[early_stopping],\n", - " verbose=1\n", - ")\n", - "\n", - "# Evaluate the model on the test set\n", - "y_pred = model.predict(X_test).flatten()\n", - "mse = mean_squared_error(y_test, y_pred)\n", - "mae = mean_absolute_error(y_test, y_pred)\n", - "\n", - "print(f\"Neural Network MSE: {mse:.2f}\")\n", - "print(f\"Neural Network MAE: {mae:.2f}\")\n", - "\n", - "# Optionally, view a few predictions\n", - "predictions = pd.DataFrame({\"Actual\": y_test, \"Predicted\": y_pred})\n", - "print(predictions.head())" - ] - }, - { - "cell_type": "code", - "execution_count": 88, - "id": "2ef77983-c1ad-438c-9551-72f6aa89d773", - "metadata": {}, - "outputs": [ - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - }, - { - "data": { - "image/png": "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", - "text/plain": [ - "
" - ] - }, - "metadata": {}, - "output_type": "display_data" - } - ], - "source": [ - "# Visualization for Actual vs Predicted\n", - "plt.figure(figsize=(12, 6))\n", - "plt.plot(y_test, label=\"Actual\", color=\"blue\")\n", - "plt.plot(y_pred, label=\"Predicted\", color=\"red\", linestyle=\"dashed\")\n", - "plt.title(\"Actual vs Predicted Close Prices\")\n", - "plt.xlabel(\"Time Steps\")\n", - "plt.ylabel(\"Close Price\")\n", - "plt.legend()\n", - "plt.show()\n", - "\n", - "# Visualization of Training and Validation Loss\n", - "plt.figure(figsize=(10, 5))\n", - "plt.plot(history.history['loss'], label='Training Loss')\n", - "plt.plot(history.history['val_loss'], label='Validation Loss')\n", - "plt.title(\"Training and Validation Loss Over Epochs\")\n", - "plt.xlabel(\"Epochs\")\n", - "plt.ylabel(\"Loss (MSE)\")\n", - "plt.legend()\n", - "plt.show()" - ] - } - ], - "metadata": { - "kernelspec": { - "display_name": "Python 3 (ipykernel)", - "language": "python", - "name": "python3" - }, - "language_info": { - "codemirror_mode": { - "name": "ipython", - "version": 3 - }, - "file_extension": ".py", - "mimetype": "text/x-python", - "name": "python", - "nbconvert_exporter": "python", - "pygments_lexer": "ipython3", - "version": "3.12.3" - } - }, - "nbformat": 4, - "nbformat_minor": 5 -} diff --git a/src/griffin-stuff/API/Trading_Bot_Development_Strategy (1).docx b/src/griffin-stuff/API/Trading_Bot_Development_Strategy (1).docx deleted file mode 100644 index cb1035354c609a1e4629c69a22a922a835eb2035..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 41835 zcmWIWW@Zs#U|`^2a4nCITpP4>;#5WkhEr?|48jZy4AIW{c_pcNCGjDZ1*yfcdKI}j zTc`T^-*ymad;eE-n!cSvwKKoysz(JH1v(R@T-MM2Y7u!u|E&+x&+lifIzul8w6p0= zoHO%#t=#QRw)<+fsYN@xyp4^W6TVo9d3nv^yNmSRrI|l0evZ^8N7l@0v^dRqTG==!pA% z<+5+?nb3w4{B5S8_8WtCJN8dIX>4i!zwni_Qtj6TEzvhEUHoeIxWwCPG&&rq{_&CT z_p?OSTM|Eb8rOaI__KG})6&~bi;{hGrfEk1ohrLTaR0=_-_^gkm#>y|u{-$Z`trL^ z?N5DITJ&4QH;w@{7Aw`f9eyw{Fx+EeVBlt8V2CeD%_-K`1CeVdo%Lci6li#F=XyeQ zdET~AmnApjTweGtu{$v1%)4cIbso!Z?b{pZ5*s0QQ11Er$H{5eW-pe$p}j1pIXo%y zYSY5&2P5ule%!J<{tm}9&OMEa6(NsWB6DVSe^uslNLBl8mKLU9r+rWOXtd_$cNZPk zIl07Z<$Ca~XjwT~@C$c|MtEth?&W;8d^ewMrd`X|?tGF{-u1(NTZuODnzC99Um)Uha*-#_8{)w)WqsX^^0|{v`yDnAaQ$B6dc`-a^RobltEU#@t1d4Ok;WT4 zznPf&=A52qcJ}_`CAm=xw9Dk~TOa>iV0YJi>vrbk)=91M?GxWOrgU*fivGKuBzEWb zJXbBtwNqOTy+82l@+Y-)tK?G$x&x(nxHimV_MB}P>vOHcO8LkpQFu0zr+okNG1GAN6i<=mj&oQ% zcX?dW*>`r)i_P`_-Tz76kkS5O{*guERXSgs=H+tde}4{rl17b}FSl=tzh-1$$YEh% z5I~NX#DW5Fw1k|_yKNw_=eIUXY)MxpyJMilq^4_*7oG1)ZJ0D?`;}8lUK3e%8a}G4 z{n;UE*{1)s(|WV@?=7c(Uw2Gn-{O}MDth|oq7wmyce%E%?>heM_Sxr3#Vo5>SBi!m z6^t#6KYwcNRW@DDl9}NF>rPcGZB9D#QI5T!*kAj6$t^yIdt0iP?e;6Ex$q~|!Szkq z#veyMObyt2t4qFaN_i=`AU!u)+~%^f!)rO6f=`!uejMR?BA&eDfkWMeuWkja z=f2ap%^=>j<@IKT;7u#TJ3jTkov=T`@%REuQ|7m!L}Rxr=tL8&7Xo?e$1+jiT+kW9ll?WPc~gOKT@slWT)J+t-y=*xZJoc;5& z^#7}0-^imTjPQc%7X4xj3^)8B8MZvXC`CUdKe;qFHLnDeGPcJ2ExPSC>Hq(ENnY!0 z1EabYr5rz*rJQ?m+T((lsqJPdUW zN~x+VH+dv1TfTh%?&aaLzT+hh6i-R^t;tqaIpa9Bfj{=0uAzwaJ2-+f5L zh&OdjaOp1V;I!}Z>HojJ{i1TkapsCG0u2*pP0kdX^UkMB$=cG!dV)Efo^<4Bjo8hu@o2dg3*ciP|8{QTv}L5me17Fen~ zJa}T)`8Q#2-Wt~{e?J<3@}00HSWo7p1b!&$j!| zcb&<~xH#RhdiK504GA3E8lP5NxijmhQA(bgQ8Sb3!IQTOR~1|KGzx0yc_@{iOPJec zHRs~}p8q>f{XFk$*gbW|ygR3VlssowJ$Xv@t<1dI>WJQl^`G@*C);0cy+gC(H{Bp8F>cWf_r10Usb33)u5T`~2wwMB4Y-Pf=ZhWKSt&P9e#p8x+o|NqnQY8S@f zhE1NnaRIy*Mk`NQur=~L6g+WbpF)nX?5Y_aB1>JA3%eZ^Vg;tV)bQ|n6>+*Giuwf2 zlUbUufkor8Xc3QC&)W{J42fPhE%k|MI&M3CIwlvExNEq6T%)bF>iH~Cst?pf1X zm8PwfIbvQNd(2ApvxNKn$a6)B8!VEI3ay(p9OVu?VmQ=x_F|WgJNND?uXFO>k4&9h zVDNmB_#)f*#tmQYcK&Q$vyXA&R0H9;;_Q2FvN)VRa;h%4J%w-Pgr5vPxG7@Bt2vR{ibJ>fHn zzo~Ko<4gzjTUmXw0h>2^iAH8|+}H4OWD)ND_}ug8n@WDQqlwz*7;D)W(@$Bf4%)Ap zYwBhMWHg-V5-jvT@9OeLp=;XzT$;p~)z+R< zCvhWPHON%;r}iR+umv_3HBv2ho>F)>Pm=e>-Yns3|E_Y(V&GBvsLzg_*%{=95%y`fGdim*juN-gHzdzb$J6m{G+v6=d7g8rq zZ7t;UYvp=!`1c+b-RmpgzGZ2$yI5H9p^Nce#0T|9Z#z6w7Fs(=3Ni6?2ncxw>^!r~ zh`;mj!gzb z*hK+a3iXMeixo5XPu~#nBJhJ5?52xr=0`1O4%Khy5zI+hCc>T+GyRZ-P_N)4*}2iTC!FwiFc6$F zZK4Yohq97<;T8J@9efX&_8$|P$TxqAqJH%u_E&y$^b)TfI~934AjbV!-=55guYWkY z&6(ZyPBmwfLe7bG?zx9l|E)|DIWD+ZIkYwDwaNdaX{<8I(>X5;NB~sT9D$@w*BbZ*E1e(xE)n;Z^EN{AsNn}C+Tc+Iej;yOCjlO z(}f%ID+FFY=;PTrmH9ApV!XjMn~mFQeW%%Ue?8+q@vKkJ(>;9m&O5hQ&$bNZo?OC` z(ABBa(RX1>&~&zADjPrU7iL{nz1{4HIsr^%_jFfKIJR+|uX zmF2pDh285zZ9)~(e^$)YiPT|!o_^iptj59umS=jrs_hFRcj&y!3y4y?*>h~k%}+A| z`#g-LPU*a2H9WpO@7nC^e@fr>Wa=4j+Fp0+{I}}UGR5NBGnenx3M##vvm#gd?R7&R zlNjp;t_hPe92*;pZpOWq>)f{X2p7kSaETY*$CYn=-F+u|k%@SbzYiOWU-<#$<~M$C z=WzaEzpl^5+)-dCQ6li8KPKz)+T{I>-Rhs$GF-0bTX^pf)0g-6rY*?uJD+trll!L` z!-hTT4$*Z#BreS1nBUdEUubzMDm%iAb# zQ{ugCEdl;tZf|%wvHIT6h>P#nKYD!VUESZW{`cjUhI05e$j*MH``mr;&cAcJ77H_R zNx%5(7A})`f6A5|&Rv)E%2b$()-li4&N*9bvC7rvcTc+f3dR0kMrS5yZByDdbCG`5 zna8HGj-E4@KT>}7=_i}>%fQfUbzNVt)GaUwxY^%ff7<2mFN-VTD^+g)$TeKwXpz0k z*m1&PLAIOARyr;53GJJEJaEZT)-OgZ_br|ByLY`jvv}k8!s&bhr8C=h_p4mk^u?~r zh9mIuqRq~YEb1%Q@f;8+m(wome)QX9MquJ1-i^6yr}OT3mpJ3osvFm4|9`c3{VT?) z8-tc4ZvJdGz3y3a=T*Vz^!}W~TQ>>nJXd)lZFwg=F`8k!(Y_4#Rd>VY1acMM;`w}Q z@upN0gPrYH)Lq*y-n6v(x9j@x|L)eYAHR0*-}{D3T6#aPf4|r>JI_}SgH$Fp&!4)> z*({~|^YzemUmWfQ<~|M!R-AFFyY!QurtFiFsq+F~vdrA-y~s1V(<#J#w!L-svrN5P z7gt&zW@Rvwxjy{Bo106fp1ZOrrnR^-^&xkkxAcl{AK%h66gbaTWW=6tb8=G$9% zE*$+8dE>NTI1f|t{hMo_oWEx4)OV!1V(V_DJ3m=B|E!(ydW~>^*5p~2`**L`+<)_J zS8PPg*4;|IcC&ZRtjo!+@n3r3*4pxr{5!9I*>#8(JkN;nR`y)iAF}hK5q~~Ia%f4` z4@=knV&?EAd-v2@?=kP{VpOp;RXI2}{?_5s>5tU9HE(#FY^+?Xl(~O>|K0<-pC|m8 z;HL0%u}8k6So-VY&_n(S+p!oh2!P_zV7|pDf+jUEV*X(^^noYntLJ9B1X>_ zeoilYz%_f5WX|f+(+B37*)i?%UZOvHx%Y!j<<5!uu2!|nk~Y^E{QRKA`#t-elFOEb z9Nu=P9rs*GID2R3r&V9&i@xgrDU@IS>Yc&eu*J@^W!4q_$q$yV56|tt_PYG$+w)KB z%}iC4A~$E|icFZEtUATz#hKI!37am~*z&o(zt-}rzgW8Rj!nX|6BioRzMlU*;(Y1W zHTT|z+wR!+@y6}TOV=H4ox_mKo67t@L~F5Ysa(;1wdA`U6)W~@FFgC$vmktH&e@kD zYr}mP$Zmh4_w1Sle|2x3b2{%h+UYMcw+<0rl|Jkua$gRFG`kkNczQVPUT^mDtbq2jxf{ACGXf$ zuDL8?EBj-#Qd(zLFF)H%7s-OpN=bzgPmXzK@*(}=EdaYacm4xnD=_ShRV?VFmu2SN{ri*wzu@l@rgp}h(RVHx&015(?VYvl-a66S=4+3ZBs&#s4GXI5+`TKr!K}!q^Z$IU zXQJ1wb0@9&lyLhPk2=5a&s*Yi7gs747et$z_op-d;?p_!%rJLPMNrKDT|%F}G5Ky) z>NtCF>!WEmvb%!Ax-YQ0OA7RRU!Qd%*|X=~a)J1%V(S)9>QPyo%imr!N7LxRy3c;962&sL zJnAPmnL0SUklJ&%c-i`>yoX;O3$=78=gPN4Z9d97`MpL`hGE1XzdeN)Tlh0-udga+ zD5$gvvWQ-BBFRIzc%IN9xbfZ zJ}tG{t^aFN=N$7$iSqN7LLD0mVpo~o{1Nc8=z3w^ZQt43!hQKO3Z8Imd^@phT}y38cEj9Y#ccH@E1M#2*Ji#EJmGQg_UU@1 zo}%cODF4psdyh)oj`ylq8h&N|y=5o17`@H6I&k>Lv+`(x&6dnfdoILUSa>{%eXO%c z{nnK-;kU6l4{vXaKVQ1^R>m>Yt}{1_qo=NB-k;Z8*qsy%gW`rF*{thj?|)eci3c ziq?NQaqUHs;+u--cc!mxpE$kU)HGf7ad1Y~pLzf8PHV78R9hJp`nTeI<=UuuxTY1|f+kLuC*(Kc{ zUs<{?+1GqiPHw`&ZJ*OcvTEjcS@Y~J72T$6h9tYn#V65 zE7en`5+f_~hbKru?s>%HQ){Q0LxN&O${-FKR4?wVB2-fGtnF{Oj;<&IDn@2%^_R@F=>T&b}- z$;V3miRs~etBln9wC}KP-nAs*{na1J;j4M7x4koPJ2%1Xb+k;&maq4GQs%9+l26Ic z5It(v5qR&h9O3;jcO5% z10_PN!ZWta3}=&f)4GbcOsg!+*r3*qNCoC+%z@6k$epx&hPfozkjlxdo6E- z(Y-UJ-(3F8*|DKu_T3P1xm70THO? zv~q95SH3~ z2YJ6w1wWONe*O4F>GO`W9i4SvNh@A`QToE$er#^uDnvrE$d^9P{Js^&FF3mGvo zFq~##V31*8U;vL^Lnh^r$FTLl6LeD?{hJLW+TQ=;I@CH@WLaLf4?ghQ9M*J5kY(eIs|m5nrm6{# z-ao(a^+$?K{a^MtuCy&zWt^8iZT#aK(<>Cu683!#^I_v337MxJr(H{Ca4FyKJl)x7 z>4rJM51BtLH|}<{P~9{uxFx$)@@%$6)|ZP@j#T+^mQ=Gd**~`HSuXR)QZ^zx@N~HQ zYuiH~W(YMZsOO3m-&zs&Ea8r~#jP(F*RxpuY`&zA8rJMr*3T@~XJ8NtWo6)pg>`XB zWlm}_XrAD0Z0zJmZkuZN?_aby{?C)#2`voA&*d>pDw9d9-aBvOdt+bY%Wp1R5YOm; z8(*)g=pY{&6dLBWYSJd_L-WJ}G@pjgog9+-_wW4rJvr~|;^q5h|9yAY{P2F>eLJJy zhy7lD_TA^*`E`FKYOH@gdsh+lZr-%pfAjvnn{9vnc>cei{r`XdzW%;%x&8Lnk7IxQ zy!&zY^sBpnKfnC)>+bgU$FC~?ZnF88`+wKHR{gqL@^RCae>i{mdHeLI4^Q8Iy?yoH z)2}~msl1b8zT94Z&A|hE{o5=%dN%$4lDRpm`uEk_r>n}l?Ee=;yE95&yk?`SrtfTkYfyr`O$_cf2|>uk3nan_uFB>1SVUuzzPi_n(-ZPm%k_fQQH1 zn}5XX-~PRO`qlFL_4nTLpFaNf*WFr+3xDmM&f2~WzxV6%r`&2Qw|cocJC%C1`lJ83 zzwN&K@!jk4(~q+@|CI3JWv{r|lDv2R+rHVqKTQ4lsbZF_`ROm+bw#>We=Vz&KTlr$ zShKWpU*)Yk_AXBS^{x8<_TAaHd;6NxQ;!k_G z|NXiduJSb=^WG}{KeGPk{-lHJPhVA=w)sy_cp-M+4}J7jl_ z^;==Z^_>&GY~W{I+Wt<+>U{fV{`U3j?Eh|vuZ^^?wP@Afv%5dFRlNT1zvqY3>N);? zx%;7^guZ??OXS;5G-u3^&NA^wY*>P+CTfgVG z`|}_Dc=CT^)PA>l())HryP5se{B^*r>bs$;`oABypOu+E=eAt=@#z2E9WrX)k~}Np z^cPF)mk#5a^@wBBjK^NQiL3b^uHZKel0OtGpR~sAh)ZpUW=+TCy3(?JUyi@cJgUwW z7xJ^S>2gTjq6f-T=WFb6d8*DO7x1&Q>7cfK;D-FC>Rius{vOGa{2%+{jK>XjyF9cdyYAZg-)}GfD)?Ub_qgY`&-QQM z+sz4d+ZucI%7??xK7W4s?r`xO6YbWWGE2jI{~vx2J`uj5aKf+G!IWBT-%^t(qix~rRKF;6% zccs*fs4ks68RhEO#T?!RnO;};R<}M-&1a0XI2r9A;P**Ppk?wN))Tr;pTs6KP1wV_WC2Jr<7Bjh7)+Ao zp-MiZBUJsmtx~FbEc%t)Ck}buV}hCe)@0Hi)+Gum`HY%itx=!ECbUl8!-}F6Vin6u z_4iqd6=)W`4^%!G?Z5?g#k4)FDGNcafLaOF2Y1DUs!!|0nx^kzO@Rvu_Z3y)D_?B-$#^hx%*~c%)0sRQ8xOM`|oZkt2}OUQ+L8vjmL#Gam znZF>WIAC~z)KW%icM6Pp>=c(C2UZ-b2_%mshIjg4gwfDZBTyjR6FIQSS zO!asiAf4F*b-|?OQ*|>(&1enDy_+gs+p2e_j4p@k)bn=ibM! z%s$@!mcRT&Y~7FAYSAfWGrGQF* zrd8+RQX%%so-&N;I|X%fYmZ&Knr`;||F(yY%InyB4Qw;6@bVXZ`M98GdPJY>oc)dh z2b>=>9L$edcYnu>y2oXQ^WA>^@bz4md9V9ejJ{NI)x{msaZjWkx}H<`b+-DV=9Qyo z+wJle>A4j1h|emB&1ib*G&yg_g$Q}^pa(I`Y|o_>eq5UJ+B?`MepcDx>1w}z2qjq^ zD%#@XIrZ~Yp3Q2r${qc_EA;*5IeMqP{{He6$$lqokIR;N?f2f*eLpDl@Y=EummiyH zF0;?uf8X`rU(=Q4^WT4Y+5hzK)86i{WxpSP`t$Cu_W!lFzb?Q0^XkKdLlu7u>#g3# z|Cry;{pAm%+$QtS1+^Ue_Lb-B_rG@jRRt=@#CG1r|bPq$Fl6TeO;zb|HxfD$5vqGQ>Dn=9H;l`Za;ZuHLu$F9ajrv{8(6H zUCQ$c)f-oP<*9Dz*e$b8uV}W+y8NV)Ub)B}Kc7C!@JSEz)mho`;P-}knMX~!Hw$0g zdOhpjp6lWUm#(I6xYWbycv^A3#sB2Zx^*At?OEa86!7!f>IXW90~xa|c-B08@T{pv zwDRAgS%RDDj+YDDPHs9DmUg_}Wo7=|&!61RnqFV+#p;@UlTTrVdErlGtvHv{&)KgA z?|cre z(9!5imzUMX-TbiUPHkV`v|aw!u08f!e9>5}IQx>>-ec+8bUd{$@3ys+`kLlNOo2QFb>=JhY!1WRxR|?i6}I$}gF0@_5OW9$^(@zg-E(%FJea-YSmDII`>UsKsx&;4babQ)-!9)({2}a`;GM$a(jFb2R`&-JHnOL1++;3_-d1k+ z)GmMWlT9%#hBv0NE?AP8>L~yEmaEDCgzMr{d&md# zRp-t)J$pNAw!^2|iJ_lQ2<*%O8*Ae)%wrBNnC;RlLE!taI9gNF&eXV)+ zfw{eE@0bWTahF`uW?|6c!D^!-cX=j);NhqG&r zz2|6uzMon7kh)Cz+8<)!HIH{6S)j&t>}-g={5ttxTP$V#7Z$#sxvoO-?#;L;hW#m1 z+a^g{?>?ZlbJMD;!k2;%YZG5 zcXO7^nuoD>!*4iTah}jt#dG=JhwWb`sLUyrsEFzZ{`-vTD*N_^6qVif3I8N{KEHvx1k5qdYNe#X7?5EX)s@K_9N3K zj=d98IrmC3`~6X9QQv;kkk@Zh`0uWJ0isJf?nPLCm3+~=H!;5Wi&@$7X?qjY6?S{S z$hDe(-aYtUbCYnz(+w^6O6ELf-NTsMCqBXTMX!$ci?cEXkACelPihjjF5bHR1#hrp z*Ne3k$IOKHZsfmnO=fQ|Q@Yi>jlvh!s$6IMVkWuc+?hwV8WxS=eBLj7k6XqakiPJ& zf-$n?-ib{u_j(xB;~x0_;S`=@F>`gTYyJwZAeVe)!!Kq=f_rZs?lhjeS@;64O*z9} z%g;BH9)6M&nR7t9UG>Yd$1?g4cq>*fUX?0XrF>eb%9wpl#6O)AdJnYD&D`sjKLcdU zI*F=cUzhxSZ}r&b+`6r+U&^qT_3_Hr1x9PR{U1y{EL3&5uuuF0+ZM|>2eF0bn{*bM zCm-;9@Uw6F{eYSCjeqyt3lLT5y=TF0@Am6~+YUx0@H1PnMf-J^1x}y+!xt zgl=A|{i~-+^Fb@qya4^Pid8#8_O6(o)1G22dgb)%+$nB*KOA#wsS5o)`NiEk98bPp zF4oaQQoXHs(~tG*rz)qVf%1{#I@}{cFy<0A=5WnZq z9k4#jUS{v?%*H+QqDo2yzdlRhfJj4=;V?x;MpI<+kVId-bwr z{xVyn_p`9>oGyBfrD&yN=ZP6R=hd%R{IAVQi?Ptsp_1dcnQLOQp&t_1rAn?6d!c`t?QE=Px!lvfpVnFKGJn*~@(wdAN#Xr@8Q#@@2+J zpWn7S$kcgtPY=_p-PflpBpABo#)x?YZ{7L-oxhD#>m04)OXp3>z3sDT-!2K4{nraG z&n>uoeb&nYKewMar_e(Ns?Yc$xv%ZLZty>+kXK`Kjgg{4En}>y}$y0tE z4&prgv<&28?Q z|5iocG8(>w`ktYB z22{kH)L2^fGvteb<*uEs4zY)#*S-o&S1;0Hx^$CUgzx50$LmT`qKuk9X7R9DCAxsD zO8IkOvu)Y;GjBqDyBpy4C@H|~DVOr!#ld0U+uqf{Iy>0Bt})Qj)!|{G>f$MW{EHT~ zpS*Zhm+6(zy{yoFEB>^FN(+(%1CCcLC|IDhpzZtp{*xY?A6(wX;jwM`d`IU?D+2-PSjC)U~{m#BQw(i!op_i2ue!8s=@lT$=dOe4S-SW=!R|8p|%)L5C zkLl3)tYu0Gh1?c%uSVEvtx;d_V#lW0-(KcqYE?R!FQn)2Bu|JzCmH~lV^8L>*9p3i#IFK2t><#1D7rb+o*yuO-z+bLDe zH0k`yhsC8LjIoVM51xKV)?!i;VSM=_3+}m)7b|kLn9BCNNX#oK?rzxCvOVgmvxdz| zjvIbBL(|1k?L*CR=@@rE@svAfutZ*aAj_Y^zgsG+e}33g!}Wecg~^AXPj^2lZ7pup zuU^0MPr3F_zD4)sSrPk2f-GK6Oc-EH&6Cb;RDXrgWcvq{3C7i|ZG< zRy?`)YuEOox4av9*ko6&;!=`1(Czr;G|z)CThlZuojq4~ri(3WDXlP?GezaJ@T!w1 zQztTQ6mgkmH19kc_q)WY5yI>)Tq&;_PB8k&2^cDDQ_}DF;T8Ua%QekE4d4v`ZgOcR-BI4LAK=(xoR zY@D%fgGASvGNsLW0u!ufXjZ3$&yVD>+{M`Er{i{CSblbrVbxsA6{_8ua<#{k4L>k{ z=r}5}>7mJtGtcJC>T5P@by(8CH;3g!;~WQ00p=vbPfZ<1OInWxq-+zOS=!ie=cLe0 zWA|T3Zj8KorORk9pYNqfM@7!Rd$>{J2dCYUq^{T5>fMdcH*9{$0rt197u$(O9WSVr;Hq)|gnN>)xG&Y|Fa^0KGY94=fdZp%CHE%__&!1G3eIn(h-(yl{E8b-G}&+Kw=ShqcW+v{v36PG>o zRW;aqD=hnHl2-ZG!{(o@|ElWVx!k3L(XZwu3tkF=Nd~EZP}Kpy)1UuqKK7`XMMTc zV%}Njy~*j3QzW+`ccu9O zwQZg`lW(M551Mo)N&E9G0pa^cjO;%8KK5yoTYtnNJ5aaK@6nvQY?4PTzh2&J^nA^~ zq~N`=mLS0+mb<4-Qc=rhj&>7%e#FSG&<~=?5>r!-Uj9n+>FVb^=S;RpzaBK{NJ?%( zbN-__5y#ZeWzIQy;qg?9blp{|=RD_ZzOB$~vAk|q__H}v-5DcZxibnpR{@D0-;(ubEVcV@Zn)BaQvvQmp0@_? zS1>=ClL2wz222-De?G@Pr|(bN){EU-yjkvNCwNbDerh#Aoo{NMRd zD&q*-y3FcClJH~sroZn>dj7lv*U;zxzx$W^^Zwt5e|35ec|5Xe@!0%0ZD!p?|8=hG z?cMWJ>WWHjmg=21_m@sJiJn~BXlZ-)+p3F?4GNYe_m|3x`4wH)%{pT_VeNM-ckA3d z!;ZRQgD*AhFSEDKHNAeUm&wpQbnzeKZvpQo`75x9w>+_WtHwVyYb!VRyHkmr9TI#l zEjo8{H7=#S`6AgTv7+3>wT+duH9iF=C6Hh-hUM4yuCTwYjgPRcGVTzSFw5Tcx$$K z1IxYa$qAf#Iw!xrK6PXfFaO6=iJa3U_`E?@M}V!KCb6R2CIa00KKFH7-UgP>V5^m* z#ikv&y-^Ts^)$Vi5hb{+{;ulx!1vFePtQ)Xd`o|%kz2-a@xA0vsX?(`_sW?7b zR@q^}^80&>c@|}EdS)(Xb=C8-%Hw$%v)oL450ov-+E+MB%{H|3%2VFunQy;mUYfJ) z>UxDmnF_PQ?rEP6dOy|GWG%Ca?}D4Xf9_llW0w_p@v>asusLJanxDPxex;X8PVP{j zwM@`gc*ES4v-0OG^YjmuEv>wq8TGpMb;Y~^s7WtJk7*q`lJ5(T$iUVMFT8ZXR`yUcy{kPMrB zf3})6$dkUyGuM7My*78*)pbx$M&Am0uj>o(q~#irCyz&hJo!GCkS7mkM*Xh6XSzJ| zt9D1mECXNAEZLE_A0+m!xNxKJ4v)5R=VOcJ7t?+3=!u`tv*$nX)GkwUf&Qu`OC44n zxa3r1xhBqV|5fc{0o&APr-%Of$2Ix%@z24{eWwH^xXyL8&A69z?xvH3L%@y0YRxag zFC0FmE4Qb)VzGlpz|}85c{s1kwi7nfVEI#FC#|zXS7qY`Mi(v?d50$q0!%JE9Lfh+ zJr}W5I@%uO%v!5%-6N79=N!@aGIjGun-53!uJPT{P_;U%qVaaENw@feyY8I&BGofl zAOD=O+4Ejfo9&8ZDF?0kmrWWV0H`H{`f!yZlN4YNo~*Pd zAm@;{Pg*gbZ^4X2{;aR;6PU7sxhF8cWa1QHKEz$%IH|%hF2MgoQ&e~d-(%Jp#|7dN zs&3lM;I~|@8j0jMo=Hc!KO9qN5$}q3_7h`1pVf4js^qWI9g)Hd6Ua9Ibg40j6$zLdkNXo2_p2`up5rhYz}b8KD<%!Lwh8u8CL z;_q2xvdIN7&$eRpJ?iXu!Tae1ffYfP8Vu7z`Z$>sn-+U1^<_PO&Acb0=JwJcu7%*98{E$vld^W3nU757#juTb95@Gn6lyLB4W;{dw>@VKI}kU)y{7+%+(+Bo z9)}wKqk8x5$QOo*ygc^fR{rXR=YEC+a#h4^Xuh((^zzL3qvBPsdc8Y-mNxvwL!Rn%}lpTYi@ zHFola-!|P2c&2WkvEr!f*9{Z;#I7zrBo)!or+s_A1} z{+?&fvlMurS>NoY@T@&)`il03<9elrm#ezZtp6KU{_2ds=+cJct@=#sVqY&a-v24B zXxCH` zc%P;JbUd)GVAH%wGv+fsm^LF`yw~9jfBVngJqOa6tJ9+LC;r@ee@n>RpDP!MeAtn| zzA_$c)wT`Os-h#nR-LQa7O~?Ehpgkw`3Dv(sI9wt*y-EP(0;Dl&(*Egy|LTBs_^%r zKeO@;Z=bWXH=n^}uWD$|d_dQ5zn)a%4Ee6uJu~04Z>xyPFPihT2 z`LAc2Ie+O{VN%m2(4wgUS%Z~Dc~C$>ibgfg;DLJz*42PeOnaIXU#phxM!`Xj?X}@UlJk zEbK*p`RIOG_}rm(Z-ZCuiet-?=C8cmze+l@Y?kEDu-7^Js-N10$cJnHm@0oMZMFO* zH&%Nu!H^HHD-N!oy7)rL!Qao=ZAJeGsC~WD=zLRRLDT$w_wr9oxi0;DdDHxrtoC1I zR;(8fILLqXM#u-_(qDn^4lKE@IPFRF+RZ=TiT(+wjamC3d}Dd|-yA=!AAwy9n&Pi> z*X7P>Sjl}4KoK`F!KADn`*i0lmJg_QA{L zpUTxK=Reu}`k2l;2iE<3zrwC_*R8Ny7xtm}C|8}H=${o2mgjdZXp(>Ym5sZu@wxS! ze1nMNbAu21uV1jdN&e6hxt`_g4nMQk6a8cI4HlW$!e*27S#jfYPJv?O3!38DYgfN! zwGWY>uKA<#^x^69X2*jUo)X_%AikeT=S?sBS&?s(Zbnq9;%aggb}!oQQ2AAU~A zeCov~^<-C^j(4Z?nK?I1o(P#V=d?0fsjhkzbfBWj;jYl$ zVHS0czCt&>#sa2?u`Ca#C9~^QPh>p!!7O3N+rB2v>e=FqAH|rKFjjwI?=oL{Ca$3{ zCROXe$43*dvRd%jt(*0rV*Z02A2`ccj&t^v3-;}v_B6swoq?lod%?V>Nlpo$6CRRed7F6bPg1LNSwrS{5Q+-*7MU9E&K}hbQj$Dcz5B^yiNI= zi>wZQoW*NYwtSjyQ4jaQkLzxr8PH#?b)fJ=pg+Wb-TQ@Q)GhA0TI8{RD>(A+*yY5h zB~6b_rBg&KE`Rj7qrh^!ce{K2gNoo&Z&+oHhc081(U&_G(D3*{vjj)qcg{ZU?`f(~ z7fAJ}y_+G-z;Zk~!rE?*#pRtl969==^ZWlD__%D!O>T?6)jq5iaXC|C8Xi9|I|;Gg zdbuFPdOKO`$Bg<69DU~e$7^rL^4u`hUxuMw(n9X|`{mm!9vCo+T=?+aLi3nKnB z`9vLfWa%(z$r=--OwLyA()#mnr~1s(+8Vh@XYXgWwE>})S*-j~Mpv7U>*Tg|UHUQm zGEZpcoA##ol1Zz7=6yN6gyUuQ^J8ldugv`7tD1KEt=+t@7dOX7{H;=C>ARU1w=;t& zq2*LNzv%_*T&t&Z);Qu7A+W^vY{@X^ZxbbGjT4T2*S-I!sTpW}G?-gCLp*}9x?bDRFeynkHR;-+$ex>Xtvb3_7HynU zR2+QhY0=)CHH^%{@|RCaW%T6OMe5W{tj?W3`RuXJ2R^svPmP#qqpfXb_H?6BMSXBm zs@evlMn}apK4PtldcD^8B!`Ll)SQ1_dQdm5EGhNKDXRmgjwNR&ZoGcL;J3Pv_GFLs zmQSBCy)N1r7j~_ov2*In6ke0CXZ_D6TUIYt%y!C|Jhk#*j!m-up|rbN8;vT%!rtVc zPEP$`w*APd36YWO%0ttbiWaQVxjx^_c*R+-y%~p@B4!>~apY7K1(z1h|@ zH#wDQqxYIMi`V2l>328#UE`1@?VFS;^wM|uq9k5|ze$V5>!#49E$Aj`pHf4sj3OsIN8yEI^@wlb4 zC4*W4zNyn~O8O3-Fse#2xSMEj?$Es5d)B?Qn6PR;j~dLX=F9^JjwFDrYL3(qWA3lr zCkgWA!86T2FC+^)Ut4?WyVs{RYi>x+I)3T_@1oABikmmZT;G0%<%v^dU6fbg4S4 zX>P6>qj!A|`|3JoVeQK&r7Eg7#GLgZLN^@KwDm9s#oGE9v@93@77G(O&}e&&sN#@yu8O`%*#sdp1oi_4$z-MG+wvwa4# zji*mZ85nFdn!Bd)*r}$S&(GYPYZxJ%G<%KC4E?k@#g`(td#&+lnd+);^l@dxAFU~M zv1Mh#e_wrJ^zAkOyRP(+rtQ9yS*&|~Ypc^XFBg6`?cUKvX=~=}UdZ!E^NrR1d%a3U z&$RYiT{Bfw(djs&)a78pCcDZ${k9I<#q|lxHhtvZHZk&m$iI$zD`LwQ2lxcsJz!Bg zO?gkXVPi{OL(4PkOF}*hGaL*W6PmWT+TP~2kYy}k=DK%Bp4Itd^jcx9DJ55ZoLMWp zeVe%UHgM@~@l;vi5YV`Q$&vLZb6m54&W8Py5}vLyTj-QfS)8v|x4P7+by8S<<_d=Q zpWOl$$_Fg`c`vrMnJe+g58=+*`CFY^@9?HBa7xTJa%64d+U3A{{rldD1?#eewSH_) zc(GAZ;&a&R7A{rWsh0v6?*()&aI#8kt`vwsC#XlV=gFVib*fap!{h3W12^ z>H~I5TXs6_D_b#P_rg!UGAyDTS}&MH*L1Ame!eQ~y4IN*`wd6)xB9IJ*!0_mSM)*j zc}}e_Oj=tu%-xY6aIj46UQeuf>Nbm&FjvG@a6nwaGVgssRPUyu(AC+ZHua}6ADr3~ zTC(E9`oqsU`2`<^FFbKTEhXUElx6$gY4TSF-fdCQs&02^SdvpD96wRwKgT`O1mRtG zMD)cqs_Y8?@>;gYn=bdew^(gLrk?%BEB9tQ&3AXr{$n8NvcKsc`-`OwY-|GgB3!Ji z%$?od2i@oKywR}1aqbO=+}#1yFA_k-Q`zKM=N>ixnjfj$lCXPmg4UxMu?IMKZFHo( zX7oN+lobDV$gGg*)mkm3)=7IOWG-R2&+X~aIKgSIhhy&cfNGC}@@LgQZ~RbO&ZK$e z{~0O8)|JV-&7{Q)x381x6_Kil*ps}t`DVqjxP-Hx4w+4a8nhN{5EtB_%>mUO3;!j* z`|8Z~)qb;^N8%el-RLKYyBokJJdU(IKfAL!akdBJ*D%qEPAC0Nm;^PvZ&6v|;IS}l z$$~B6D{Pkp{4c!pOR*K?k0l0Nb6(EaE8TkY@u7?*7Y-bLBr6&5`{j!&xhLy7M7+$L zmQ7@;QC3l8b>b>jV%@sd>8)bxuYQZG{16|oR$CMpd8Pi%ygHAC|I)8cky6y5sjsQ{O@yD6pe%P&EeI<6*x=F9Rw6?#E zuXmW=AatVN<3!8bx*3`dk`s0s7a!;A`0=m)Uq$7<-?cmMt~Go8{XWlsvmej$cGlgn zS$A#z{rb=E;~(d%-@o(rTld$;k199E?fd`i{L^={|Nngd|JT3Q<-7mRKY#je?5E#f z-|b$%p8x&(`QLtDe}8Z9ZCe}Nc(L94_pXuKHv8_g%m4EEXHPGm?LR+^|8#l0`R(K7 z=MpR|{_Ma0_F3Z2+p5Q&b$@-|efRImozr>av-P{%KmRXfYxex~J*zhJ@wYy+^`}dV zK3z`xz3GN}eV)Dg{u_0b)ql-0KbLNduZ^)?|E+#m?$7CIb@lcCx5d_%XFa}sx^~I_ zpL#nR-`CYw+Qsi#x$9fS;^qG>|0kU;o_u_M%k1{;)1Ow=@7P;@`)P5};n`o`Wqz%C zYku?Eq{;rz-L&K_K5vMRZ=ZK-LHq5a*Pqm{q&o%0iA8Mz*o|W^wF6{N~{rTqk>#tu6%Ub()$G@UX?MOD4TUNb+-!7JAnSXwm z`Ei!dg6Zw=>;CMymm@Aef3me#)gL|ACx7ezU%wl>$;LKM?$FcaM{j?wn=WT5bNk;W zTUl#4D|ws$rGKpNt@*xt_xk&~W&EqJu0Q!#b?yE&LeBPQuI;{BTa>o*#@qDWSCy)b ztYp2<+bf=}c?w<4YZv4s6{rdCn-?u(@cID4?U!HtU`unq{Xy?Cg z_vJTzpK1BU(D3KKx{ZIo9e#7`(D(O7cMEVGxz&k z+m`Ypc=h?!A|~f<*#G~yetXFovHah@|McQEmrpdGDW}Qv>fz4sCzZZE+##+iu4Dgq zOXZJEHGekv`|a6pAK}mU@AvWeZL80kMc%o8-+%e?X8CFQtL4M``%i!U{rln1Ltihy zD*X9>-n7^C@2B-&fBaus{@MSPN88WduBxwD;LUFPUT(3y+_d#C7tM*;6u0E?pWf?aD!$b9V&Kn}64#?|r&C$K&5Oc^CYsD{GeDH~;Q| z*7xb=FCP59$t&@ruFUw`SNrM+nSc9!KJ?%*-M{YJ`O~+)=6Yy47Ud{UX;alPP^2^Rmv;S?lpBMBkI{9=VM~gG;Nfvd!k@xQ2 z-@f{IRQ2LXf7c#6q<_A<{psarKMQy6{ayO=c-*Rq^UcfU`(#$s9pJ2by8P<%<$q%j zwaZuP{8D8;nEGV-)x7xpg-;gl&?(?u7teH8iu-?r-MV{m=WLH2oLDpcP+xHV@!*Y< z`MAy>?wQhCVAHuiNkz71p1P#A^uvltotp%=-|M)eC91bA{>QUN5qaCro1dTkcKP=7 zccG~|=Fj!Ze#p7y-HPuiSgV`7({g@v%lkBb{!Zsp%&AweMjV*(YLiYvS6Z!@fyd?m zZV8o$jjU@<>fZey^>5#;xt!a7?09dqXZts?yOa0?lZ^iV-gfi)($XN2+{CW|$Kd5mX&L^56L*w#?xeI%Np z-SLoa!(pk1Vh0j?3%DPAIJI&c17C1RG()@QE!~F03U9>@Bu*&cez3u92WtV`2$^df z+Zq1OlHRZS2&4$68>Fb>Y%ZhBvK1gxwt$QWn}VzW*%a)qkTBiOS}T(<@!Ud`1~I&g~3**-&K%^7FVK7`Sc)nRA!Pg69&* z9ApJZe%|1=h4o6b(mf^%9-%x&u-Evqi&lI7;`{dbQ{V5_r;1Ob)&8vaoV~PoK^`b?CcrkF=+K#p$|xWfo<>-dXLe`Tsoo(yQ!mrfZKsHs@YmewttW-;au) z9B+%NU;e!_cb}-u+}-znztmgDzHN6(*SvLdJdbwTmp*>YE7x=7{>`UxezpcF_s@Qn z-nRO0e|O57e>?T|KK%NrzxL(H=U0E7{`utQmp_>s&&CwVc6@FP-ucdA&bb=FnSZ{f z?yUK1xpRMcYqI^?Yo}wsJ@|U~WdCXZ*}0c$?>%Wace#AZbklbQU!Ud9uXFi%`ThCn z);sSwy;DAQ=la_F(`#?<`r&_c``VW$FXcxU@845*dzRd}`rEVmlHc`LzWXSD>F@8H*aoA_3JCnTnS~X4!+4? z*{sa+?9Z=sh7@V8+9t`p(#hfA-1 z_A7hz*52;Ur|xxS2U9)JF~D<(7hLO<^FKVI=*XI=Hr9e3s>OVOe zc?;#-X0Cnm{H*2ku$yb=o&S7Jd@Ji8HSSl5Tl_5p_MVpCZW&)1qdoj2RgEL~MpEnAeo)^u+6xxAE8j$i>}Kif2!qB9n@$9#SZ zIy$!O_{vduQt%Oz;#H2olGA*%TN9D+Ph@;<7xBuO(~W*m%RV;x$EVX)jD+#_$nk1af-aYvavGSW`0HQ$GjN6 zV}}bj+1@LO`Q_qw_gclv9eb4OiY9Is*SYs*&h+^fOSzw&xVSALjxo+|N0MDeaXDZ&mzA<%X_u1Hw8S%ym zjO7+ujmsZTJ#_iwmbHbJH$s}Gx(wNxID}&`eQ$8vim+ z@j+FJ_SIJk>dR)u|N3}(!3M{xpNwYi<1*{7mlL{Vyj#<@`b_xw`_4BlOY-*GNgVoo zD|MfK`sPFn&Yz{579OtZpPN>D$~L{!|B`ON^o5V#-T1^8(&%?m`qg!Hu}b&3UmIlJ zFEnSK-z{DGt8S6vLeu;FrE`7xH%LA7u|MP$C$Zmd%L(nrYt)}q=xXh? zAJw;0C#`)J;K)s-#5*GEUIs3IDz|plt>tp2xoy|KthgQ3|F5PZ zb?G8!fANT`a~DtBy*cZY{Q8W#Q`W)v=S~xzJ4^0s^t+3zH0xhoJhpsKRDWOe>x-3_ zzGVFk6JoV2dHGam^R6wYN>eUC4by-cS`p?{g*F=I3IlqmTe03jejRDU%s={#IoRD;jPA(#{2ThPHwY5 zmlHfSoUi?6LGjn6k8@7!`V@3rwqnn8zfI0@l`CKOR>+_Bt_ZG+-uZ2=M~L}Z)o_AnM`m3&vU~Q=KxuP_f|c9X7ZMW}Ctu+&F`4z)q2SKnUxnXK z9>4qk_zvmG-<<@6A6NPGoHR6KQ=8?jtjN*fX{(_2q+!MvW`VZUQ~{S`TjV5zi&A`k zH}X_Zmf0Zt?&qTkKWFt$YrpKNw0!o*P48E*iC1kZi%AXH$iG)IVoj0Jl<`_>QtZ z3ejKSG{ZVv=|j2RPpgAgH`cD59o+inb!bWLyGvhx>fOwlFZ?S0-onoZGagJW$@n+<)%HVm z9$X)0-P@;|=5piit({sw148Qb)~C<8w98Ml(!(fqK8O3U-29YNyCO`dp04{gMOTQwn&7wb*jdSxJ9BCt%~p1_Q;QDCZn^$+Rpzrf z%L3OnXp>+ds=LoztvMNqn>KeQMpA+bio=Y_HZj9{Fuv`m!qQG9T-6_px<6CvFS4ow|3sV#a*fsIe8N2 z(T}(Ik1H*?Qnf#=kdJqNj{e%3sNC}3ENEGD46?gEFLU6` z_k6hfk%_fUG~XQl)@R@N%@eM2U7x+Sd~el0%k{h$Soci1=a8tbyZ_1dN#1-1w94Lo ze4tTcxxKcw(N zo<;Qj)7v*a53c*3_^xky*wn}G+KiQFuHXCl9Z&j#$g2OM94UvFHhMdAMdmIEdc&u) zcxpj*!@+Ha9M_j!Ut?8r*rJPT&8`zyAKZzs;bD2_n^`B^RvkR?*%Ik@XRj33Ue~U^ zy?FmS-sjN&Ikfks>H5A$Es^n&k<-GT z_aEB3u|$I7&g@G*%{A)|VB#KfD)V#l!NhSN*}g6yyG~OM4}c z-Yc<5e$RG&f#3f-vriuQp5_<1f5mR|Id2c|ZLD_Yi_E4{rq0V!5q8qgk9%n(*1dU z`JE!u9p+;E%#GELxp$b~fia=%=;C0QZG_RNqaMQ@izYN zZ0~q2Q?>?rZ?xXf6Ln}WBirp`3U_A5eST%pvuDGM4XaOFeQ+%H5UT z+vZlAwLiGiv}5{brdyjb4W@2CKhK)CfA;qN>o4D}UaMMl-SMvW!fRUv^4b*h+>$kq zHP-sxP%o=s;D}~Wd|UM~WsTqZ#XjM@@jBI!k54W+sye-jnwIArt z-zyv!DI7PmW?F`<^Xy9T*oy41s_YreZRZ-d?KNdsoSLwrDtqe9dzQ%}A|=JzTU+wR%=CbjqZRPp(BZ7FM~&#m(4;?G~}nZLGSdyP_jMYeB6 z_9xEArAwOgEgsr@UBsP#)+?X4_`n6r3){A{C2Zm0(aeAQ{`S@6sng3W*1wyRDm?js zrL)z)X_r6BOX-I^zi|0azSOLC@637LJA2Jv|JBQ_{=OmV?<~H1zkg3ZT5on+K8;i0 z(y{5Bv2)&rb-y|JqD05`|CG2Bf1V!Bn{O$zMlSrQV&eY^~eWqkS)3R(-tj@x+Z^Z=)Swd{q?be<`Q%uH}%&XUAXflQg3p zbgIwXSTyHA?Cy|dj`NZjH^?$bA5go0XzA3Nd9VClmQJ}kSK)2CzH_O%^D66&d-7x1 zSf=w`S$;~lX2Q?h?_b|)8U$&5ogJL$vO4F6AEOM%KV82>k$}TIZD06b%9k8CwslFc zA_If7vB43M%NKYbYwWsl>iXKVa&P?%7&xvt6-zxe2y57L_5g!Jp8bi^b`CA?t~6&B z5mi?~r4w;&lNPm25lQ{|au*MWo}5P4tmq#CN*iK@H!NQjo7r(-&Ze>`)?;O>^pspK zJQ8uc8h0k4g+oAzL3Tr-?#qx4h65LPcnrUqbUB-+h?ks{f9uzQY^4Ar%*xj{(w$jE zO5cF3wEEU|Wn((j$~kNTN)8=S7b3 zN>3qwK%Tib%FSMGrpyKbCXwLI@#M!6Gg$vIS~(ba&K?j5_CIVf$7OQ&+BH?Lp7xcgyD+e1Ewt<{lF)Xzm)60cuwDMiE@u|y zdjbbJ6q$sbS>g<4pZCrR>*uO#a?vtuytY)RlXHvyWH+A$l{=$Y4zamxp7K%Bp@Tt@ z$=ZbRo#f)`qpM;!pUM{Gn0aAmPNt~D&+BF_99BlHmf3<4#;u0U(K8Nm2uA!lv4Hv8 zI@P&cf(#O~8WmT}v%08rPCaG+tW$ERQFAqlg<hnq5 zC(vjcEMdH(RkBB*@n}DfRl@tIODhE!8x~4RB)!U%af`Fq0gA9X>=EWCtMuder^2t_ zK0VC*IOEQr39|AQamD4kUc6Ik)MjDdG?hm{X2H+8gbDw&A7$!w32@xKT}y+RLCH7%JH5ze|oNH zp4hI`d0BdDi0;nmJAW$tJs$IH)A1hBgFZL3qFy>X^)8)za_zGt^W(lfE_rr-{hx!Y z%`dH%XWx8q&yu30Z2RL*&N&mdJ|!S&b@(Rrr*}_Bwbb+f-zWG#uFlJ3--#_7&Y0}K zAmMy(PlXZBy8a2g6Axa_KX&=}oi}Iiv>d;`>$v=`L>X6A?znFnHFck>1mo&&9qyZP zeg0n${c|_RPsQKTYNF%K72-ZONXx$nFon9)2}{^!2oV z=P~h_V$)_m6PHSUal6Jqsv>9o{lN6B55Fw;)SWqd?~HH7-?i0$1Lst%@jKZx-Ii

zJQ_f4{#sT+TC!I8H0 zDgLKkPv6S1UOui~Id1=-DH$~rGZSZK)-5z_+b60e_1ru2@az0Ta9eg=mp}GUW!4m( zikpi+{5t#Ewc_t_BPXZ0|4Uy9?5Uf)=ikQ^I}Y1#Y^JO3ELdVzt6Cf+|8=Frl%Grd zkIuVO{&$kt!>@;5=O1}JeZEkyzLBnu@mYP7V;8nNO*~Xn>QnL8x8m=N*?VWs{_yMe z_g`z*SLPntmRKK{+P(MRq)U5K{?^j+Ew4+U6Q7edABg39021 zZ8myp74qh!DC4v(IQ;X9?Tex zei}YU*~43Wj@TVhVR6rE9*p6=dUdi4A+kwT_U9(ut6%=xdd`ynTGw8`nR$gfX4(Es z-)#o}c3wMu<=_HFiHt3A4~ze9wFsN)v!!+(gNVig2VYs?RRu2t!gjR8Oxw#-B>BYX zMC#v|hest6`n@=v3^t1|6F8$VCi>%O9E%l z2u@d+b(NVXMQV{`K*E%aW8v{P7p_^)b5W;S(#05L#;=`+x(yG=iEe7mNYnRGoaJb; zx%lU6h9y>t4`0Y;cBDP)@D@1xijmET_acvmftSf+ZU1)x)70BDrtRff1U3WeiB}HJ zCLm9oRhWg}6TL@VHhb@l16i&x%WwhbMIP-va%s008@&=Z`=(VBfH$k>*yu*mXYBwe{}auGoC<{Uytd)e&p!e0P8O`}g!C``K^#H>VVO=4Z|{ z_xp?ie#b9;_J7%WCf@GNx0#0zzAC)hHGc(f+7W48=ft3*jcpsvW!u-J7f9V} z7OlRyDlkMO{?lz|=XaJ44m?br@=*$_D}Swjuw7POv--{Tt5yI1-RnJD?y}W(0@M7b zTCDX?woWO1;8?&dKFh6uyVSzJwjj78xa#1e9C;OSmF5(tPp_sy}=u~M2* zN@Km!<%jRmPrC_C5qHRF`Q%o3NZ{tpBcI&gF5Dv}?>YBV*!0BH!Es$uIScl*a{Dgc z(<&vgGxezIf<1HIHh*gRWX5C?5a+t=@ZIy%dQ~*^8!qrX{km8~#_;FDZ0)C4zk9rz zvVB*F*St?bApwj}bz?H5p6aG-IH#*E$N2PVj{FqyDIi~b(lU~T_@a`rB`q$eV*}ebV-S;B#XKwyAy}!}V%;Q<;qztQhQa1UFSy_IGe=8qIuJdZn z2~0U~L?Ag=>SbBbe`%xh+3vk^VSgox+uU!SnDE-~eS2n{)sqt!>uCKm#p?q(JE7c2sKVp~^bXufD;@_E>e5ucTIE)3-JT{eeZI*E^YC1J3Yl%#&!{N>g zL5*%YI~R&h+4Ned_r?UrC5;}7StI^Rlx|LyntEl+j8{vW`4+Z)+_lE^W{tJ()hz}* z0@)S{S0+Uq{``q2^;~*~sX(^JC*SU3S?5iWrzd5roNkK%nYz(UM|ZvGmQAmdK&Gx~ zjDVVYRBGyyS71{&woN>EX_n2GxnkK-5^No&G6GqiYaXr$dk_$&-4MaNv296Elf`wz zC49{~JKbiocbLxM&=t)-B6KR+QG2)EwGLj%BT^lvQ4Lr71I?q?1ZuK3t&FpJA&@Ph z)o3ajvP<`xi9Ot{f$7H=uG!4Iku8bWvsb=wU9tzOVq{S-ldC|sgt8-aYZg<-Y);=FAUi-S&Z1@A4npm@z--~NruzJ)J8I!mmLF^QMGB}ypULhFd- z6wyZt*BsU;uJlv?cyhbs*7qqVO}$<_u6gK^QF%^kp;BDfg;VRgO_QA2HnJu4iY0ta zzH(M9JH*Q%Y1MQ-t0PivUJ1TePvv!+p5t-!*H}i-T7TID@ zEy<}X18VGT2@$b03^7u7=AFvpHNjc6I_S%U1B(N9v<0v{`{3_kj*E2VOkN8m)B5W{LoU)l8)GCkf!@>W;77lg#}h!zJ; z-50`hYD!k%Ms-(FP93fw*U0V#UB6a)HfW`NaCQ?FeI~JD(ZmqLp7h5v&)Xi%SjG&l3+uP;9!rVsHQIAL^V}`uT53{$_BL=ydlsah}lrN z%3~t87$^wN?(UlWQ{l^v?guNX1gBJm2OgbsWp|erTk5^gkd5l!*Dc!Rda+zo73|)} ztRR!`?(X_j%oB3aW0i$;>*9{TZS2`iHyuPx?Iu};t~uv2eR-G2%oh%#o2OrVl0TcR zx$vaYi*Ws>pq~LDXFO9Yf(U_?Mz%s8Kx6L)l0f?0rTZ@-GasHI){ z!XI6!m#mCk`jV0@_1xv>%o0=I@+bH7hVXyaR99%#KMvXxvZc~d=+spkiFEd&w;sA# zMLmH}ST8=2PtWIF{_2TiudVrw)<;Wj2Sz*OiRGyTh8>u>s5?>Yf|SRdJMl|C>}VAi z5HUZNz3$&5&4%T*AI|Xq|5(HrF6(`wG>rGO7t=f8H}j{8#9U-xRFIap_YTt1eqr=y z=1ZAVXG#RJ6s}BBztXFIc!5dvlvN?hmv)5|r8a&`-K5EFqCed(@t4a+ z!pfg*^WU##-rO4arnlwr=PusVV-X$}6ADg!p26qFygntckgLa0>FQdC*FiILS#M4E zOwqouC+1#MBJ-lmte-p%hhDmbMG33UsSs)nJaIRELV?1=n6`Tb$7O{0CNOa>fB&HH zkOI$-pAD=+eYTYcJym7$jP&#$e)wF=l)LdK*v^S3?=C#^gpZS5$glF!2?Y)}Rn0GV zn35-)(AK)JD?};~!qsjac3J-4tyLX2i zdZ_mJlMtW%{c!d)?!`BxIolV1n&}+*lShb8WzMG*P0&E@&F)hx4oIjJJ>=M;HK3#PNDpB@r4Q<%lwex5;Pj$cT)ap0lOLpS=D-}%eR$u4v*^#yPF8O4VTniC4D z_~sRexo$jto?T}`!JC!&p0l~zr}v!pw)j&sUqRgOGsIe@mClh9MZnft&3wG_{Zt*T zx<*5fDM6wcv8#$!Is7(i4Pu=&eRa21j>8rnt@(@|7IRlrWi4tEcHbB-v@``XoUJPk zWk_vu7V)u5XlzOLzndAv**+U;FT;$p{YlSNW%^_+L{{c6n|b-IhS}ZAmY-f&p0!b{ zN!iNh$Q@MGzJOQqtn-DIJ3-p1E+reEnlt@ftZlj@JT<2BsOMtVz{Pi;s?S{RtCP94 zd)DKLGcWF4=UF9g6&6vd6j8vl+4vgE%C*JD(*&O9`bS!6&)C>@CW7nYrE8kEkCd!p zuzGkXq3P4o>iCzM9;_*$xgTD2n5`CGI7um||7ZR4lPtCk>djx|mv7Epz{ejV-IujD zjc57ITWQms8XkV>doe-zLP4GTmtUKdl=wO<6dpBnIB!?tTPS@@b>$7gVg_sHGY1{A zLRZ|LJAZSBmBOnFvD3MV#R}WS)I#bX@%{QfZQdb~iR|76T{oCDTFW@jA9IxXwEJ#{ z9mDkwyBUW2uRP}ObNKTx=!oP;_9OBF^$Sdzel#FlTG}Ii-IMKbmmhBWWQ|K^@CZX_1~QvA9HMy zr@v3xdAGx!;d+NX#3Gx{Y3Dw&r^Gn@Vf@%`eMEkjRQ-p=8fU^|E%dMayrkG+-!NbC zUjoa{S!|95N$IVs9t_<-m>==a*j#ot66CfGjLSRh75*hyxe3-YEWIG_#C^Enm|o10 zqTPb^4C@5z4U+3t9+$Ui{NZe7(E!p3Qhj0R4`+$?$r*=YR@feM>ixm|L-G;7?&Q@c z4}ACPs;~TGv&u{P&>Yu4jK$aHS%U2|dmsV|3!mc}kN6je+)|V|a?n7gU^j|alpB8> z<^ZV@sCW3YaP>!a^;t)MPN{g_YV|*q1rlNsj_dX$n9bj8uP6KQ^9rqAj$*No>}MjY zpXof}@36no5DfB?f$ZWA`-Y$w>;D^YxeHpi97FX>P2-Q|ez;FA{&1FYi}PGyasGo~ zJ%hdTpMxAlqWz#Q-2U1+xzMG8iJ;i9%svnD&PR6Xf-aEOWyvBR*#o%#3h+JKJjuTz zBG3F0{{hoS{Ap+OFaB`eaii{M=GNProB3Ye+JHUO=d9wl&g3zDtF~;?m#Ul&tj*W8TF{9KHTA`-WPxU+a48hk4XT;R}T$GW^R zl=1M2_zedPvLxIh_y713bZ^5kd%OJ;%1d=0XdDjQb|moIqmu>+;cMDf^Mp&T+oV&l z=5D*}`QF-Z)zb^6Nj|=jnSS%D`^PDBSy?k z@iNcyn`X-^?Id_@wq1Uc(k|?0JSWZKZ0(KC+T$N(wx_r@T)v~fyz-V*Qr?3_2|f(ic7FS*8HXV$I@vJ8wR0xy-X%(yv&)EJ^=UJKyd@ z8&BWrd}vaBBUA2o@bS5yu^VRVx7_me#Y35GmvdS#p9u~S2N`Q|w)SS{-5aufH{RD& zu9++0SKL1LS>PMrwWH#HMgoev^Vwdma)j5#)?vlfy%x&okqH~{3 zdNs$~&)6~5&*E&stl6crD!ry3$ZWgJvwWw%^~aJoz3i#lOmm*qS)QHv`aJua2sUA! zf?2y?n{01nwKy9qXD9HPv!EMna^F-uj5^z$+tGqIi2l(`2C0H*(DdxG|u#SE6cUVaEt7LvsZeL z&T`89bL-nfd%w7+zg_rM^r|^(G-myKT%*9Xv~I7DrvCRoPm2ma=^6ZATa)`IOvH8K zt5Zp?g)N&7NQ6u})M^{-$@O(#(c0h_DoSs?HB^<(I%~h2876eh?~I>1+t!x+FlXJy zsHFK(kIwADKArn~krI~<+11I>CM# zUc640nlTx3ih?wZCn)pAIC>WpEG(32 z?2y)3C)lAOU?kP(@mol$(c|+4j;dvX7J)LGTR;8TJ>>#Niz(AP*OPKW?7a)Lr5er4 znfI@$Sg&Tq*Zbhlg(#DzLleZl{rvn;%(ThOfW6m&uV`(KelCN{b~l6h=h+htXS-V- z-sGsv@hCxJ(^8ph$`9Xi8Q)u{v2u0fvN-`$uD28g?YLm^bV)#zvb)ruJD>EXUQ~{M zYjpkBiVM3!-1r;6r5aUnnD9Sc5`W55&J3P98R{l83@b19LOeCo2Kmh-90l1dHrzd?SF zG`=aBg`eEt$@J7E!t!C7Hi3Hc42wwr|%P+cK5b+VZO`%@Qq$P@{& zFBIQUyl|7mG~QF5GH23_w-p_#QqfS))l|1%TlQOO=aYiYT`rS+=Wunlgef(i&hlIt zoP5J&>TatNo{tX8CR!KYoNa4dd%Y%aUU_6io6OUFo{Pe37~bwndl~Tg+Qe;+1`i6g zDqYorwm02dp%U79KJcgC4~4(Sch21rqN-h5XIl58|9fN1KlY%+g>zbYE_-XPmS~zL zrpKcDHaXN%AQ%E;-5h;P1wd1nlcr4E7IfL5!|mgQ{}X4aE-&4>Xw#D^ivmPWWIlA7 zaM@1m9it^H7+y^kc@n~G+54zv-lllQgj@}?372nhHXA&sUy^n*GuEE9~o;OFXQ6e+%_)8U+k@X5@Fl0{Fxs1-|}&^)nnQs9JO*(X=}7Zn`YyYPmC z-{MFco|{a3pTQ2Oe6ab+6eUg@r~}lVvRR5so-8z(Tg~L^@1wH3HT}fI%d;;zD=b%7 z?&ap_SK9LRaY4`|&-SegBOLwQ6JsX1H2m~9JL&SSklm3*UK7*BpKK{mJdr8?d0x|; zuAP#}_D{pN;bJh>N;qpwFR4VR**iZ&U`y`I{9MO;OzQbeAFxPLln@X5<-?nW~a z=@XfsJT6aaDmmkx)iCE@?e3~Iic8zQuk2jGtv74`gbCf7)MmZAb55N{>21;9J=4!z zyvSfYZK3rad41V`k85(ed^)b^1li1-|1~xv>Qdahd0IW8k^XZv>KzV<@tiizF7SWp zx64X$$$yJ z*6Qk&zH7r=SGUCpLOna>Wppl0@Y^!Y)z9{N`OTi$GuN&@ud}Il-=(ukhbzqHZQL)H zdRA0AV#zUgrG$y4nUgl@Oj7a{ib#3rwBf|Q4IVG2h-}+@;zIwOhQCSLuMZwn;(h6o z9r@(Od@DWw_(s=lm%?{NKC$6BEcriXW20+UjS7=k9KUb;q9Js zRyNVPd(Qdi&6C3SNAmNDWqPPbomO(+XfC@sn8#r9=5I4yw_Ob16`3Nyet6RP^oZi4 zY+@@vPUMJ4iG8$Cr+bNq4U1Uh4gsA_`Z_^YC!62u7=1XrL1nfK+(ApM=TroL+pT;i zhM_iXV(yoz-@YrkKh#K?UXg(0JIN-H@0R33gTu zlC``3y?tzFnxLCCld(!b?dj8+9b3bePVUs)exP&b1TL1Q?3I=qf2=eY_WM%2_Sn)V zT@g_MZf_HXSRIYt+&S@c!!@P1F+G;LTTO~`=Xz>g&yI5X{#f^mWnuK>r8?Pshs*@k zr6isywB8I#-DCbQey`$=uYP6kMK?Y^^ZM=K2R%jKgTmK~Jzw&iXUa5HiDvO5GyNxY z*%kb0UVQzAxWsPljfJhx5--g8b%xhkeWF;^f>yO(1#?*SFYoT1FiV8zne&;nir$`I zYO3;YF3UPTXO);8Xg$Y*p?l)%J3`zX(yTp}ReauTEOXyp?p2eS`Bb&nmo+=c__s^f z9j~W%&n;TedRq75k}V60UA?|+n{X)We?L0#ZQADUaj%clN~|S$WB!W$lSFqgW%-`p-@njb==L{1h0fr$nXLZn zPVTfdII-c_oVsS|z`NIv1%0`}b3i}kLmZPr%#-sT-OHPscsgFs{5ikLRD7D@s<4pk zmG50zC;wc1Q0@5jZtYD*|30{Sy$)@DU^egUGTDtXx0DZ>_#JwpGKuA6Ng)@fesg3^ zHCN}lNiVN{5;5wX^yXCxo4NVJPlr8c&rjdGgLjgOU297BldM^#dtX`|3f}%(ZmYfJ zEZ&OjjS71%OyLR3v-X?gws3y_$`2KcJM6HmYQOXV%Ndp@%Hq>8#wqKc#x-%I-VSoOcS` z19VrNx4!#jW&U})tseW+c;r0}-*po|@+13Q2m6s5(#|Jqe(+1)w=XM6w<=%LDjK*d zwPqvdb7wd6XlX0s?s(CCtIxV`?6lI!+OXp2EdHQfNrAhqzic{jy6?-fIJ38@8>b)L z`{vSVz3RwEshyK^3;&p|dy_uLxxVDn))b$23xBOXUN2C3`i}2+-`u6~C-1VEy-m9( z#Cp2yzOsC(_6s=%;~PsqByBt^YNyV;IOFI^?zk+0vvWTC%{Hz6-MjAaoK01)7b;Er z@%O{A9Wy)YpFX;`aFSB}XJ-9h&z~;R?G9}2_Ic=?vh>@H_q<}qB1$UT7q>5WlRN4t z#OnEjYrEq+i<<|6F2)r~vD|I=wLjzfeMZ#W{vzsWFask4!!||+1`$|pFHcPhM#}KL z7Y=3}HsE2nP`&TalS#pkQ+f=YCK*Y{?w;bfjlV!f)o`nB^3Jorz8goHN$mcjrub5E z<-K~Tvw9Ca7v^p<(l=rYd$i!ql@$5+?RS%pX>eF`^oq@T!!_mY?Im%mSby<3b@U!l zN?K{O$@95D%u&mel06G#WN$>sT$C0}dZnJ%Vn6q;-=adL6>*uO>;A0Mkh|3Tal3{^ zpVuMdQ=1v3uT{QtbQjb;d+K8UwmXY-U$8$9WD4AK{6+qGj@0^@pZ>|GO!?wr(PksJ zId@A?;U1l9CtggEjNI!dw)cgpNoLC>vqYV`fQ;VLw?DS#|7>vQnY-hFaA_egFZ)tO zvFv}(Z|8nuMGdjvmJ5oLnHU&;b1^Uo!9px8Kd&StF)1fiuOc_+?zDq>w+v*Cy|3-k zm_ElVXYy7NL9JtLTTb&yS2GC9Y!JIW@vXzoOTW)q=ky!A-QcFkC6KnZjdr!9|Yxe#WaBmX6 zKU47Ooo)J2m(w(MgiYHyceNUKP4$}6J0-EAJ)DPEsU;oimEbzL!EAcb6C)jawU4O^ z5k01|7dl?e>}J@psaU$QqnpE7YC^b@hr~mxz4xc881*%+HtfB6Dz~cRcF2iD9ao1b zM$u`3&yE{B&6a-gR>DxqW1-86>@?$_T|wH#)+{eOT(=!%k(d}a=^5P@H z&YhcOW|_9^U)?b_C9kC+J%-|44vYN@p4?vWBzgDin+GMlUB6VBX=Ny;naoVl1 z=jF%s{OMI8U@TK`jQ}_>0xHmDeH2Dy(!v%)xslA;4Sp;mi8}7DP zq?Q@#_mZHG6kb>Mz;>F=yO z*=>9;;w3b1c4-6_Jv-`qOO^e&(odV3{qaE`uD))UWBlDw==1km)YG6!UWNP1_`V%~ zyS2Lc-_dJQ514-Osa$b?yH)d1_Vx9#JGj^Q&vz_zDcf-#y?cj&e&?ss8+ za(bs=s+`EDiM`%@&Fqt|DKE|1c}J&G;J4U^h3&Vl-Mz?hn}bJWY3W-ghT?y66Xo|k z$kLJ7%6WV1;klDNv}f^$TU1Z|8T{kS@;%-XH`4|0QTtl{weacZuKQk&k0ONO;{(DobD~U>oIxivT6RC@4YOMdaX7qs%_ak|I1qDucbqU zeQljDYi#gMUwk`vR)&=1tTa2{|DUQh=Pk)p%ZO?EySGM2bn8op%L+mUi=LzzdV6o$ z5_EC>;wv6=$_kbGR_KeREMk`rx;Fjwmb;uHH`tS;76@)pUB}HSXH=u&xP$B9kx5tP zoc;WKN<_nL9`;4KwK5KoNqyJ$)&6{Zd;0l3jKTb zTmN6(FU=pXf0pOZtJ%tMV)B9e^}pV1kKZ5DAmDlA_4@gLKR&woI^XX9=XQ119lUJq zo_83kJDROuFuV!y*NWewDz!;u$|r1b7Ad9CC-HCJgdSKq{bmVn|C{)I|^n0rkB=OoS!^mAvjIk14aaHAr7g~pi| zvu{l9nedOjCtOlVnf$Ia_66(F zs>J)gO?Ho}je-?!EL#*|x9a(Zs%d?#{~mSh>Yo^~$?5Ll{sPC?8EkdPTB2!J7?-&zikt*-?#gF!0K+>mc+zO%YWZ?a@e)HRI~U) z`^76}50Bh;Gip3veXG*n`h8~lisia%yjH(Ysk+|vuKx4xyBAB(N0;1|PIM|eI%UD_ znEZyAS;t51r+F_z$CayzOXv17L$&k`R|?X8DrWN$jM zSYbiTz4^MeTJbBs?0WH6WZE=7anmo~-L2$BqOWvbyL#$b>%v8!Yl9z6o3dx8)|D&Q z)R(%w{gk;fuv5a{qu=ZCjH)L;-p0Qa>|JumXYQ{jGySJ|X!>rJ%2eMhB=ls3fiLS% zn>pu;dcuq5$O-Pc;CMW8`i8otW!o07FjzVL!uKtU*6z``b*L$S#oexJ=3*s}nwz(6 zy|=w$h&Nn*{f5ktG9+`UX8jEHT8RF_luJP4>L@hwi&opCTVRg zHA&XmU7{p#yL2wIfr{`p=B1^t0Td3sobNODT%Hj!C5&Oy3{7SRr@tTUoqw-9W;tE`(fJ(jC!s!X9vCJ> zm?b_mYT0)5%)S44*S)S*Ea2e2ku%w+^~^1&NDYG-=^ft$C08Wf3(4EmV4d%GbJyyJ z#Y|?VYp!2=xa0r+y+YUSKV2CRys&Ok`-7^Bp<0Ck(e}Ez{*TVg4_H#=vprOr)%D$e zxo6B8F$?UTRB{>5R?uF&&uTsYi@r0~0`Z{@Yt+P}H(!$D=Udtv6}k4xy{zu<@0GFw zj=fm4`R6nx!EDR#_jA_FZBf$h*xIIeh39AO9*^MUWp#paetVCc_As6BGj)a)Bd@#{ zr};$=KaWSxlWZn9cg{P~oEzq4!@tt_syj!2Rb)|5I$!3I`;N*DbDf=)mMlvAc6sHS zrzy`pHr=>jn4N2RO|t!=gQRiz^yI$Cl9^9eYNherSJ6*UXPc?qm!hGiwdj~6Cvu>SvWrVeLOzs2A6u|L>R zE53ORR@av^Ffgz&GB5};Ffb&S7MJAbM&#z`XO^Vqf~&o?JsY_W8Hg}^uX8;)>Etr2 z4w1`20_uVbxz>69y2UnW&!i{Da6Dz;%uAapOT-9&g{N%rX ze&+C5p!IdDBm10n0vW;S?vbTCDp(c$nZ>nEo)mc;V<>ioC9H)hy#Mr$7YfEcCGHbyQ;10cW}*4DQdZT z6U|?Cr=DKj%fP^(&d9(ZhvKjJqSTyXgzvyt!1kWsn044dfbGFvf5j~st5VZ77Djv% zV!G*6ptyR%`E4Cbg?D*-+}pP(m|4>3p6vhc#kuwTt9Ll6^5wwu8##wEYrEw(+FwKFB-W2cLwlH^W<<@sw*>I&$D&QO?ky2Ua=FL_?+ zrgur{7iRDJcj@PT#og+`GH%MB6LXeG$Z_7=SU77>v+SA?TK2KBnf!0f#aG||3o0);3QxvJskyo0Vlv)O{_S|Tw6xD`ux#LWwH-a!aE-rdIcXfZp~p2UcF%PA0soa$Db~h&I!1i99FnzeTRC- z$_HKCTuXO}X15Bgzx~9mP%YpQt6GGdNqE4;H^;7pZ3sMA7<_f1p>pVf40q3&XI>Y5 zcIsZ}(&628M$uR z`kM7Kn>9{quTN(=$J&?C#Z~CZnSWhJ+p_k@x3fzuMa%E{-L_}FujX!cr|)g&-msWe z!hiU}A6|TU?NZrMoBV`Dd;-1h{@KS^V+2Ac2JKls-{W=MR`neoOB*)+*|l`9dAa}T zjZafFb&h(A-gY$Fw&;!VjHqpW7j!57;J&5!smLQ>=3;R*joIfetN40I&Z#;u)4$fd zf5Ke81k)+qE7i18X7YT}x$2}G<*OIiqp2c#ko#BQQM<<5S^w7Wn4_AZ`uXLP*(bPy ze(M}Fa$lRc@^y>$QO6e+R`>1Fl9s7naB+>eo2wB0`d@yf_NA;o-ZzAAI6mdiQ>Cy3c+vHazy&lh$fieYU*0qE@%|3hz7EDPU5@#B(6! zm2cFR1xKd|sGQgO`}uwSzyEw&4$vmi~(^m-z#E2mTdj>(B3Lm$9VGGe|c4pb9S!Zupl`-VV$Jdl>;`$Z`QGWdw2Ys z;Ir!o-ksUt$g5*;c#qVL13bmq2@56JvJ#w)&setJRR6#5-hXb?5^K?kYw3Gg85mj= z7#PGE7#LFWlflK5eo01YZc<)iW{zG~L2CNjuVPWBa^rIuKJd79tciN6@Ww%gSF-zXANGN&-$Vv|*ntAjhc z0Jn(5F{P}JrZyt+%pdREUBVgAGBN(#o4Ysv?wj-ZZt;1`>gRcjjxUNoeBqN&HKzd& zb83a-mo58F=WShAV6ockji^}(!;9@Z4?SeaX3F3>wP^E|q^MmvJ3a4n>-034oMBMV(4$qhOj?{K9HWcVwZ_f6dYl4v#Caem`6mFKf@+SvPz`=>t>k zqpaOsftJ$=OzKmbSPglUs`fl+XNY3?!j8yv`pcgR!2g}s%V~R zUvEyFbEN9+Bd-0w4@`c(ZkFsWy>!KS{zdE#hxm?6y*&B#{p;WUop9-qyHxuB;g{On zfYQzXKBfPt+s8hC@1K(`P4)-pA6YHlWw5c{I^u|gts^UZn%B&?ISyO z;q9|uKeK!Ctth|Xz3Sf8((n5l%hz9#UO%mh?e$g$#^$32w+s#j?Xkw68pUbxO1U`SRIK`POxCzDwfH%FXkDZeZs}!)8%c{*`(6p+mz0B z;r)Utm-rH``{rt&9S*7Kv&{I-Jui}7c-vFKhU=;gYi4pr^Udp2j$?T->uFzy??&VK ziTfBP?D<*wYJCa!Jm2MuE_PSMy)f*F?8g5=NR6Gp;8bM8);a~l>eQUKn{MQb-sCe|moV?egGX~hms}K1 zSo&WsFZ+VRcIoEMcWlXA-I_%`f)UDxC1<{nljz?0X_dmi#MFqs4_39OCR{NQc_YE4 zbGT6FVa2?J%rnd{*pr^L_ny|h#2f$QQh)r^iQCf-Kbyii?d@IVIqm0s&#V`|-M%C( ztygBMZRV#bb9OH~x2O8ro2En6YmRQ4bjhf?veftOBh?>OcjXR7zyG{`$`g;^;?u9L zwae8#xw!SUTy@RkWy|N)J-hjQUj4%2)7PfQ)urBEy?*b{YqzcMRc_CIzOLWy-sh9| zYJc1ReRKM7{8t&vsx{v$?!U}GSM&Ss_rK}&7WEGwt=}K}>uTD!L%y7fOemAO^*l@;GhnLG>S^p>9mPrdj`Ht;WNVw1{P(1?}zZ`|?ikbQtR zBa;X-Xn351!L>X-a&6GkiBlOF7*4T4x*ZHq(7*^{L_6o_m89mC#D`QCq!!15T4(46 zRH}J9{9s^UxW~l6zzxy^#SM(d7#SEKZG3Rk9$jNc8+ULfBLl-lCdlL*R3AeF<8vmY zx+gim2s{#vuKmmH+v2Yo85nX{7#IX#dKnlP8W`(YkhCWj6kv08c)@jxelZ4y8-D1P z&tPL?NeGz$ZR42tDpvQX_1pCcQBH5Aydte;t| z&%huQ%F4ixV$vZws7b{ol{u;4=}Po4NT@_TP!guUficMsMKjv*lIZ53&yym|x!eph z2Wb)oT|4^33qt$BcBuC9RFr8LbQ91gA`m8g?tq$rGA)6wAAQsqp}(ygsvkUhj5LOf zZUlP&8(~EKB&@x0bOX>kx(EZpCnFmGNf@9$Ah`R7HC3SZ9T8>)OhGmaWD2;61~v-a yMd)2EgemD$p`n7(@j};+-X28gXPSwW#ZX&^0p6@^AccYqf((f)3=Ci9fp`EJJXPZW