app.py

import gradio as gr
import pandas as pd
import numpy as np
import math
import matplotlib.pyplot as plt
from sklearn.preprocessing import MinMaxScaler
from sklearn.metrics import mean_squared_error
import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.layers import LSTM

import yfinance as yf
from statsmodels.tsa.seasonal import seasonal_decompose

def get_ans(inp):
  plt.close()
  tickers = yf.Tickers(inp)
  x = tickers.tickers[inp].history(period="15y")
  df = x
  df.reset_index(inplace=True)
  df1 = df.reset_index()['Close']
  df['Date'] = pd.to_datetime(df['Date'])
  scaler = MinMaxScaler(feature_range=(0, 1))
  df1 = scaler.fit_transform(np.array(df1).reshape(-1, 1))
  training_size = int(len(df1) * 0.65)
  test_size = len(df1) - training_size
  train_data, test_data = df1[0:training_size, :], df1[training_size:len(df1), :1]
  def create_dataset(dataset, time_step=1):
      dataX, dataY = [], []
      for i in range(len(dataset) - time_step - 1):
          a = dataset[i:(i + time_step), 0]
          dataX.append(a)
          dataY.append(dataset[i + time_step, 0])
      return np.array(dataX), np.array(dataY)
  time_step = 100
  X_train, y_train = create_dataset(train_data, time_step)
  X_test, ytest = create_dataset(test_data, time_step)

  X_train = X_train.reshape(X_train.shape[0], X_train.shape[1], 1)
  X_test = X_test.reshape(X_test.shape[0], X_test.shape[1], 1)
  model = Sequential()
  model.add(tf.keras.Input(shape=(100, 1)))  # 👈 Explicit input layer
  model.add(LSTM(50, return_sequences=True))
  model.add(LSTM(50, return_sequences=True))
  model.add(LSTM(50))
  model.add(Dense(1))
  model.compile(loss='mean_squared_error', optimizer='adam')
  model.fit(X_train,y_train,validation_data=(X_test,ytest),epochs=2,batch_size=64,verbose=1)
  train_predict=model.predict(X_train)
  test_predict=model.predict(X_test)
  train_predict=scaler.inverse_transform(train_predict)
  test_predict=scaler.inverse_transform(test_predict)
  look_back=100
  trainPredictPlot = np.empty_like(df1)
  trainPredictPlot[:, :] = np.nan
  trainPredictPlot[look_back:len(train_predict)+look_back, :] = train_predict
  # shift test predictions for plotting
  testPredictPlot = np.empty_like(df1)
  testPredictPlot[:, :] = np.nan
  testPredictPlot[len(train_predict)+(look_back*2)+1:len(df1)-1, :] = test_predict
  # plot baseline and predictions
  plt.plot(scaler.inverse_transform(df1))
  plt.plot(trainPredictPlot)
  plt.plot(testPredictPlot)

  x_input=test_data[341:].reshape(1,-1)
  resize_var = x_input.size
  temp_input=list(x_input)
  temp_input=temp_input[0].tolist()
  lst_output=[]
  n_steps=100
  i=0
  while(i<30):

      if(len(temp_input)>100):
          #print(temp_input)
          x_input=np.array(temp_input[1:])
          # print("{} day input {}".format(i,x_input))
          x_input=x_input.reshape(1,-1)
          x_input = x_input.reshape((1, x_input.size, 1))
          #print(x_input)
          yhat = model.predict(x_input, verbose=0)
          # print("{} day output {}".format(i,yhat))
          temp_input.extend(yhat[0].tolist())
          temp_input=temp_input[1:]
          #print(temp_input)
          lst_output.extend(yhat.tolist())
          i=i+1
      else:
          x_input = x_input.reshape((1, n_steps,1))
          yhat = model.predict(x_input, verbose=0)
          # print(yhat[0])
          temp_input.extend(yhat[0].tolist())
          # print(len(temp_input))
          lst_output.extend(yhat.tolist())
          i=i+1

  day_new=np.arange(1,101)
  day_pred=np.arange(101,131)

  df3=df1. tolist()
  df3.extend (lst_output)
  len_lis = len(lst_output)
  df3=pd.DataFrame(df3, columns=['Values'])
  df3['index']=range(1, len(df3) + 1)
  lst_output = pd.DataFrame(lst_output, columns=["Values"])
  lst_output['index']=range(1, len(lst_output) + 1)
  the_max = max(np.asarray(df['Open']))
  df3['Values'] = [i * the_max for i in df3['Values']]
  return plt, gr.update(visible=True,value=df, x="Date",y="Open", height=500, width=800),gr.update(visible=True,value=df[-300:], x="Date",y="Open", height=500, width=800),gr.update(visible=True,value=df[-30:], x="Date",y="Open", height=500, width=800), max(np.asarray(df['Open'])), min(np.asarray(df['Open'])), max(np.asarray(df['Open'])[-300:]), min(np.asarray(df['Open'][-300:])), max(np.asarray(df['Open'])[-30:]), min(np.asarray(df['Open'][-30:])), (max(np.asarray(df['Open']))) * (lst_output["Values"][0]), gr.update(visible=True,value=lst_output, x="index",y="Values", height=500, width=800),  gr.update(visible=True,value=df3, x="index",y="Values", height=500, width=800),  gr.update(visible=True,value=df3[-300:], x="index",y="Values", height=500, width=800)

def get_seo(inp):
  plt.close()
  time_step = 100

  tickers = yf.Tickers(inp)
  x = tickers.tickers[inp].history(period="15y")
  df = x
  df.reset_index(inplace=True)
  df1 = df.reset_index()['Close']
  df['Date'] = pd.to_datetime(df['Date'])
  scaler = MinMaxScaler(feature_range=(0, 1))
  df1 = scaler.fit_transform(np.array(df1).reshape(-1, 1))
  def create_dataset(dataset, time_step=1):
      dataX, dataY = [], []
      for i in range(len(dataset) - time_step - 1):
          a = dataset[i:(i + time_step), 0]
          dataX.append(a)
          dataY.append(dataset[i + time_step, 0])
      return np.array(dataX), np.array(dataY)
  X_train, y_train = create_dataset(df1, time_step)
  decompose_result_mult = seasonal_decompose(X_train, model="additive", period=time_step)
  trend = decompose_result_mult.trend
  seasonal = decompose_result_mult.seasonal
  residual = decompose_result_mult.resid

  z = [i[0] for i in trend]
  z = pd.DataFrame(z, columns=['Values'])
  z['index'] = range(1, len(z) + 1)

  y = [i[0] for i in seasonal]
  y = pd.DataFrame(y, columns=['Values'])
  y['index'] = range(1, len(z) + 1)

  a = [i[0] for i in residual]
  a = pd.DataFrame(a, columns=['Values'])
  a['index'] = range(1, len(a) + 1)

  return gr.update(visible=True, value=z, x='index', y='Values', height=500, width=800), gr.update(visible=True, value=y[:100], x='index', y='Values', height=500, width=800), gr.update(visible=True, value=a, x='index', y='Values', height=500, width=800)

def get_info(inp):
  tickers = yf.Ticker(inp)
  info = tickers.info
  balance = tickers.balance_sheet

  long_info= info['longBusinessSummary']
  curr_rat = info['currentRatio']
  quick_rat = info['quickRatio']
  short_rat = info['shortRatio']
  debt_eq = info['debtToEquity']
  volume = info['volume']
  market_cap = info['marketCap']
  curr_price = info['currentPrice']
  rev_per = info['revenuePerShare']

  return long_info, curr_rat, quick_rat, short_rat, debt_eq, volume, market_cap, curr_price, rev_per

with gr.Blocks() as demo:
  with gr.Row().style(equal_height=True):
    with gr.Column():
      gr.Markdown("<center><h1>Stock Analysis Tool<h1></center>")
      gr.Markdown("<center><h3>Give the Ticker of the company you want to analyse. We will provide complete insights on the given company.</h3></center>")
      gr.Markdown("<center>To get the ticker of the company, click <a href = 'https://finance.yahoo.com/lookup/'>here.</a></center>")
  with gr.Row():
    with gr.Column():
      Name_of_the_company = gr.Textbox(placeholder="eg, GOOG / MSFT / AAPL", label="TICKER of the company")
      btn = gr.Button("ANALYSE")
      gr.Markdown("<center><h2>Analysis<h2></center>")
      gr.Markdown("<center><h3>Inportant Information</h3></center>")
      info1 = gr.Textbox()
      gr.Markdown("<h4>Insightful Ratios</h4>")
      with gr.Row():
        ratio1 = gr.Textbox(label='Current Ratio')
        ratio2 = gr.Textbox(label='Quick Ratio')
        ratio3 = gr.Textbox(label='Short Ratio')
        ratio4 = gr.Textbox(label='Debt to Equity Ratio')

      gr.Markdown("<center><h3>General Information</h3></center>")
      with gr.Row():
        curr_price = gr.Textbox(label='Current Price of Stock')
        rev_per = gr.Textbox(label='Revenue per Share')
      vol = gr.Textbox(label='Volume')
      mar_cap = gr.Textbox(label='Market Cap')

      gr.Markdown("<h3>Regression Trends of Price<h3>")
      with gr.Tab("Overall Trend"):
        trend_g = gr.LinePlot(visible=False, label='Trend of stock over its lifetime', height=1000, width=1000)
      with gr.Tab("Seasonal Trends"):
        Seaso = gr.LinePlot(visible=False,label="This is for one season", height=1000, width=1000)
      with gr.Tab("Residual Variation"):
        resid = gr.LinePlot(visible=False, label="Residual Variation over time", height=1000, width=1000)


      mp = gr.Plot()
      gr.Markdown("<h3>Price over time<h3>")
      with gr.Tab("All Time"):
        mp1 = gr.LinePlot(visible=False, label="All time", height=1000, width=1000)
        with gr.Row():
          Max_all = gr.Textbox(placeholder="The Maximum price the stock has ever reached", label='Maximum of all time')
          Min_all = gr.Textbox(placeholder="The Minimum price the stock has ever reached", label="Minimum of all time")
      with gr.Tab("Past year"):
        mp2 = gr.LinePlot(visible=False, label="Last year")
        with gr.Row():
          Max_year = gr.Textbox(placeholder="The Maximum price for the last year", label='Maximum')
          Min_year = gr.Textbox(placeholder="The Minimum price for the last year", label="Minimum")
      with gr.Tab("Past few Days"):

        mp3 = gr.LinePlot(visible=False, label="Past few Days")
        with gr.Row():
          Max_rec = gr.Textbox(placeholder="The Maximum price for the last few days", label='Recent Maximum')
          Min_rec = gr.Textbox(placeholder="The Minimum price for the last few days", label="Recent Minimum")
      gr.Markdown("<center><h2>Predictive Analysis</h2></center>")
      Next_day = gr.Textbox(placeholder="Predicted price for tomorrow", label="Predicted price for Tomorrow")
      Next_plot = gr.LinePlot(visible=False)
      Next_plot_all = gr.LinePlot(visible=False)
      Next_plot_year = gr.LinePlot(visible=False)


      btn.click(get_ans, inputs=Name_of_the_company, outputs= [mp,mp1,mp2,mp3, Max_all, Min_all,Max_year, Min_year, Max_rec, Min_rec, Next_day, Next_plot, Next_plot_all, Next_plot_year])
      btn.click(get_info, inputs=Name_of_the_company, outputs=[info1, ratio1, ratio2, ratio3, ratio4, vol, mar_cap, curr_price, rev_per])
      btn.click(get_seo, inputs=Name_of_the_company, outputs=[trend_g, Seaso, resid])

demo.launch(inline=False)