Upload 2024_classification_pipeline.py

2024_classification_pipeline.py

@@ -0,0 +1,117 @@
+# -*- coding: utf-8 -*-
+pip install boto3 pandas_ta tensorflow==2.16.1 keras==3.3.3 --quiet
+
+#Import utility libraries
+import numpy as np
+import pandas as pd
+import os
+import yfinance as yf
+import pandas_ta as ta
+import boto3
+
+#Import Machine Learning Libraries
+import tensorflow as tf
+import keras
+
+#Import Charting
+import matplotlib.pyplot as plt
+import seaborn as sns
+
+print(f"TensorFlow version: {tf.__version__}")
+print(f"Keras version: {keras.__version__}")
+
+"""#Data Creation
+
+"""
+
+data = pd.read_csv("http://www.nasdaqtrader.com/dynamic/SymDir/nasdaqtraded.txt", sep='|')
+data_clean = data[data['Test Issue'] == 'N']
+symbols = data_clean['NASDAQ Symbol'].tolist()
+print('total number of symbols traded = {}'.format(len(symbols)))
+
+# Function to create features using pandas
+def create_features(orig_data: pd.DataFrame) -> pd.DataFrame:
+    data = orig_data.copy()  # Ensure we're working with a copy
+    data['Price_Diff_1'] = (data['Close'] - data['Close'].shift(1)) * 100 / data['Close'].shift(1)
+    data['Price_Diff_3'] = (data['Close'] - data['Close'].shift(3)) * 100 / data['Close'].shift(3)
+    data['Price_Diff_5'] = (data['Close'] - data['Close'].shift(5)) * 100 / data['Close'].shift(5)
+    data['MA_10'] = data['Close'].rolling(window=10).mean()
+    data['MA_50'] = data['Close'].rolling(window=50).mean()
+    data['Volume_Diff_1'] = (data['Volume'] - data['Volume'].shift(1)) * 100 / data['Volume'].shift(1)
+    data['Volume_Diff_3'] = (data['Volume'] - data['Volume'].shift(3)) * 100 / data['Volume'].shift(3)
+    data['Volume_Diff_5'] = (data['Volume'] - data['Volume'].shift(5)) * 100 / data['Volume'].shift(5)
+    data['Historical_Volatility'] = data['Close'].pct_change().rolling(window=20).std() * np.sqrt(252)
+    window = 20
+    data['MA'] = data['Close'].rolling(window=window).mean()
+    data['BB_upper'] = data['MA'] + 2 * data['Close'].rolling(window=window).std()
+    data['BB_lower'] = data['MA'] - 2 * data['Close'].rolling(window=window).std()
+
+    data['MACD'] = data['Close'].ewm(span=12, adjust=False).mean() - data['Close'].ewm(span=26, adjust=False).mean()
+
+    delta = data['Close'].diff()
+    gain = delta.clip(lower=0).rolling(window=14).mean()
+    loss = (-delta.clip(upper=0)).rolling(window=14).mean()
+    rs = gain / loss
+    data['RSI'] = 100 - (100 / (1 + rs))
+
+    low_min = data['Low'].rolling(window=14).min()
+    high_max = data['High'].rolling(window=14).max()
+    data['Stochastic_Oscillator'] = (data['Close'] - low_min) / (high_max - low_min)
+
+    # Supertrend
+    supertrend = ta.supertrend(data['High'], data['Low'], data['Close'], length=7, multiplier=3.0)
+    data['Supertrend'] = supertrend[f'SUPERTd_7_3.0']
+
+    return data.dropna()
+
+combined_data = pd.DataFrame()
+max_rows = 2000000  # Maximum number of rows
+
+for idx, symbol in enumerate(symbols):
+  print(f"Processing stock {symbol} at {idx} position with current records : {combined_data.shape}")
+  try:
+    data = yf.download(symbol, period='max')
+    if data.shape[0] == 0:
+      continue
+    data.reset_index(inplace = True)
+    data['Symbol'] = symbol
+
+    trimmed_data = data[-2000:].copy()
+
+    # Shift the DataFrame to compare consecutive rows
+    shifted_df = trimmed_data.shift(-1)
+
+    # Identify rows where 'Open', 'High', 'Low', and 'Close' are the same as the next row
+    duplicates = (trimmed_data['Open'] == shifted_df['Open']) & (trimmed_data['High'] == shifted_df['High']) & \
+                (trimmed_data['Low'] == shifted_df['Low']) & (trimmed_data['Close'] == shifted_df['Close'])
+
+    # Drop the duplicate rows
+    trimmed_data = trimmed_data[~duplicates]
+
+    cleaned_data = create_features(trimmed_data)
+    combined_data = pd.concat([combined_data, cleaned_data], ignore_index=True)
+
+    # Check if the combined DataFrame has reached the maximum number of rows
+    if combined_data.shape[0] >= max_rows:
+        print(f"Reached {max_rows} rows. Stopping.")
+        break
+  except:
+    print("In except for ticker: ", symbol)
+
+new_column_order = ['Symbol', 'Date', 'Open', 'High', 'Low', 'Close', 'Adj Close', 'Volume', 'MACD', 'RSI','Stochastic_Oscillator', 'Supertrend', 'Price_Diff_1', 'Price_Diff_3', 'Price_Diff_5', 'MA_10', 'MA_50', 'MA', 'Volume_Diff_1', 'Volume_Diff_3', 'Volume_Diff_5', 'Historical_Volatility', 'BB_upper', 'BB_lower', ]
+combined_data = combined_data[new_column_order]
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+