Update app.py

app.py

@@ -1,7 +1,5 @@
 import pickle
-import gradio as gr
 import matplotlib.pyplot as plt
-import ccxt
 import pandas as pd
 import numpy as np
 from sklearn.preprocessing import MinMaxScaler
@@ -9,35 +7,13 @@ from sklearn.impute import SimpleImputer
 from scipy import stats
 import yfinance as yf
 
-# Fetch data
-def fetch_binance_data(symbol, timeframe, limit=2000):
-    binance = ccxt.binance()
-    ohlcv = binance.fetch_ohlcv(symbol, timeframe, limit=limit)
-    df = pd.DataFrame(ohlcv, columns=['timestamp','open','high','low','close','volume'])
-    df['timestamp'] = pd.to_datetime(df['timestamp'], unit='ms')
-    return df
 
-# def fetch_yfinance_data(pair: str, period: str, interval: str) -> pd.DataFrame:
-#     """
-#     pair: e.g. "BCH/USDT" or "BTC/USDT"
-#     period: e.g. "100d"
-#     interval: e.g. "1d", "60m", "90m", "1h"
-#     """
-#     # Yahoo uses e.g. "BCH-USD" for BCH/USDT
-#     ticker = pair.replace("/USDT", "-USD")
-#     df = yf.download(ticker, period=period, interval=interval)
-#     df = df.reset_index().rename(columns={
-#         'Datetime':  'timestamp',
-#         'Date':      'timestamp',
-#         'Open':      'open',
-#         'High':      'high',
-#         'Low':       'low',
-#         'Close':     'close',
-#         'Volume':    'volume'
-#     })
-#     # ensure we have a timestamp column in datetime
-#     df['timestamp'] = pd.to_datetime(df['timestamp'])
-#     return df
+# Load pre-trained models
+with open('model_n1d_cat.pkl','rb') as f:
+    model_n1d_cat = pickle.load(f)
+with open('model_n4h_cat.pkl','rb') as f:
+    model_n4h_cat = pickle.load(f)
+
 
 def fetch_yfinance_data(pair: str, period: str, interval: str) -> pd.DataFrame:
     """
@@ -47,7 +23,7 @@ def fetch_yfinance_data(pair: str, period: str, interval: str) -> pd.DataFrame:
     """
     ticker = pair.replace("/USDT", "-USD")
     df = yf.download(ticker, period=period, interval=interval)
-
+    df.columns = df.columns.get_level_values(0)
     # bring the DateTimeIndex into a column, whatever its name was
     df = df.reset_index()
     df.rename(columns={df.columns[0]: 'timestamp'}, inplace=True)
@@ -132,8 +108,8 @@ def generate_advanced_features(d, other_data=None):
 
     if other_data is not None:
         other_data = other_data.loc[:, ~other_data.columns.duplicated()]
-        d['relative_strength']   = 1 #d['close'] / other_data['close']
-        d['relative_strength_1'] = 1 #d['close'].shift(2) / other_data['close'].shift(2)
+        d['relative_strength']   = d['close'] / other_data['close']
+        d['relative_strength_1'] = d['close'].shift(2) / other_data['close'].shift(2)
     return d.iloc[:,1:].values
 
 
@@ -146,8 +122,9 @@ def create_features_and_labels_with_advanced_features(btc, eth):
     # df['future'] = df['close'].rolling(window=5).mean().shift(-1)
     # df['trend'] = (df['future'] > df['close']).astype(int)
     # labels = df['trend'].dropna().values
+    label = btc_copy[['timestamp','close']].shift(-1)
     features = np.vstack((btc_features, eth_features))
-    return features, features #labels
+    return features, label
 
 def get_data_predict(
     btc_ori: pd.DataFrame, 
@@ -159,13 +136,12 @@ def get_data_predict(
     limit: int = 50
 ):
     period = f'{limit}d'  # last N days
-    # fetch entirely from yfinance
     btc_data_ = fetch_yfinance_data('BTC/USDT', period, timeframe)
     bch_data_ = fetch_yfinance_data(symbol,    period, timeframe)
 
     btc_data_ = remove_outliers(btc_data_, epsilon)
     bch_data_ = remove_outliers(bch_data_, epsilon)
-    # print(btc_data_.head)
+    
 
     if normalized:
         # merge with ori if you still want to include historical yf data
@@ -179,13 +155,12 @@ def get_data_predict(
     return btc_data_, bch_data_, None
 
 
-
 def predictions(model, X1, X2, name, n_steps):
     features_, labels_ = create_features_and_labels_with_advanced_features(X1, X2)
     imputer = SimpleImputer(strategy='mean')
     features_imputed = imputer.fit_transform(features_)
     y = model.predict_proba(features_imputed)[:,1]
-    return y
+    return y, labels_
 
 def plot(y, label, timeframe='1h', ma=5, n_steps=None):
     if n_steps is None:
@@ -206,11 +181,6 @@ def plot(y, label, timeframe='1h', ma=5, n_steps=None):
     plt.legend()
     return plt.gcf()
 
-# Load pre-trained models
-with open('model_n1d_cat.pkl','rb') as f:
-    model_n1d_cat = pickle.load(f)
-with open('model_n4h_cat.pkl','rb') as f:
-    model_n4h_cat = pickle.load(f)
 
 def predict_and_plot(timeframe, limit, epsilon, n_steps, ma):
     period = f'{limit}d'
@@ -219,17 +189,10 @@ def predict_and_plot(timeframe, limit, epsilon, n_steps, ma):
     bch_data = fetch_yfinance_data('BCH/USDT', period, timeframe)
     btc_data, _ = normalize(btc_data)
     bch_data, _ = normalize(bch_data)
-    # btc_data, bch_data, label = get_data_predict(
-    #     btc_ori, bch_ori,
-    #     symbol='BCH/USDT',
-    #     timeframe=timeframe,
-    #     epsilon=epsilon,
-    #     normalized=True,
-    #     limit=limit
-    # )
+
     model = model_n1d_cat if timeframe=='1d' else model_n4h_cat
-    preds = predictions(model, btc_data, bch_data, name=timeframe, n_steps=n_steps)
-    fig = plot(preds, label=label, timeframe=timeframe, ma=ma, n_steps=n_steps)
+    preds, label = predictions(model, btc_data, bch_data, name=timeframe, n_steps=n_steps)
+    fig = plot(preds, label = label, timeframe=timeframe, ma=ma, n_steps=n_steps)
     return fig
 
 interface = gr.Interface(