app.py

from flask import Flask, render_template
import pandas as pd
import matplotlib.pyplot as plt
import datetime
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error, mean_absolute_percentage_error, r2_score, mean_absolute_error
from sklearn.linear_model import LinearRegression
import os

app = Flask(__name__)

def create_lagged_features(data, lags):
    features = pd.DataFrame()
    for lag in range(1, lags + 1):
        features[f'Lag_{lag}'] = data.shift(lag)
    return features

df = pd.read_csv('./dataset/BBCA-History.csv')
df = df = df.reindex(index=df.index[::-1])

df_display = df

df['Tanggal'] = pd.to_datetime(df['Tanggal'], dayfirst=True)
df['Perubahan%'] = df['Perubahan%'].str.replace('%', '').str.replace(',', '.').astype(float)
df['Perubahan%'] = df['Perubahan%'].astype(float)
df['Vol.'] = df['Vol.'].str.replace('M', 'e6').str.replace('B', 'e9').str.replace(',', '.').astype(float)

# Calculate 30-day moving average
df['MA30'] = df['Terakhir'].rolling(window=30).mean()

# Generate the plot
plt.figure(figsize=(10, 6))
plt.plot(df['Tanggal'], df['Pembukaan'], label='Harga Pembukaan')
plt.plot(df['Tanggal'], df['Terakhir'], label='Harga Terakhir')
plt.plot(df['Tanggal'], df['MA30'], label='MA 30', linestyle='--', color='red')
plt.xlabel('Tanggal')
plt.ylabel('Harga')
plt.title('Pergerakan Harga Pembukaan dan Terakhir dengan MA 30')
plt.legend()

# Save the plot to a static file
plot_path = os.path.join('static', 'plot.png')
plt.savefig(plot_path)
plt.close()

selected_features = ['Tanggal', 'Pembukaan']
df = df[selected_features]

df.set_index('Tanggal', inplace=True)

lags = 90
lagged_features = create_lagged_features(df['Pembukaan'], lags)
df_with_lags = pd.concat([df, lagged_features], axis=1).dropna()

X = df_with_lags[[f'Lag_{i}' for i in range(1, lags + 1)]]
y = df_with_lags['Pembukaan']
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, shuffle=False)

model = LinearRegression()
model.fit(X_train, y_train)

y_pred = model.predict(X_test)

plt.figure(figsize=(10, 6))
plt.plot(y_test.index, y_test, label='Actual Values')
plt.plot(y_test.index, y_pred, label='Predicted Values', linestyle='--')
plt.xlabel('Tanggal')
plt.ylabel('Harga Pembukaan')
plt.title('Actual vs Predicted Values')
plt.legend()

# Save the plot to a static file
plot_training_path = os.path.join('static', 'plot_training.png')
plt.savefig(plot_training_path)
plt.close()

mse = mean_squared_error(y_test, y_pred)
mae = mean_absolute_error(y_test, y_pred)
mape = mean_absolute_percentage_error(y_test, y_pred)
r2 = r2_score(y_test, y_pred)

last_date = y_test.index[-1]
next_date = last_date + datetime.timedelta(days=1)

latest_lags = X.iloc[-1].values.reshape(1, -1)
next_value = model.predict(latest_lags)

@app.route('/')
def index():
    table_html = df_display.to_html(classes='table table-striped', index=False)

    return render_template('index.html', table=table_html, plot_url=plot_path)

@app.route('/predict')
def predict():

    return render_template('predict.html', plot_training_url=plot_training_path, next_date=next_date.date(), next_value=next_value[0], mse=mse, mae=mae, mape=mape, r2=r2)

if __name__ == '__main__':
    app.run(debug=True)