model.py

import datetime as dt
from datetime import date
import yfinance as yf
import pandas as pd
import plotly.graph_objs as go
import plotly.express as px
import sklearn.metrics as sm
from sklearn import metrics

# model
from sklearn.model_selection import train_test_split
import numpy as np
from sklearn.linear_model import LinearRegression
from sklearn.linear_model import Lasso
from sklearn.linear_model import BayesianRidge
from sklearn.neighbors import KNeighborsRegressor
from sklearn import preprocessing, svm

# load the data

df = yf.download("tsla", period='5y', interval="1d")
df.reset_index(inplace=True)
    

# getting only the adj close value for prediction
df=df[["Adj Close"]]
print(df.head())

# variable for predicting 'n' number of days in the future
forecast_out=10

# creating a new column to store the predicted values (dependent column)
df["Prediction"] = df[["Adj Close"]].shift(-forecast_out)

# creating the independent data set
x=np.array(df.drop(["Prediction"],1))
# removing the last 'n' rows from the independent column
x=x[:-forecast_out]

    # creating the dependent data set
y=np.array(df[['Prediction']])
    # removing the last 'n' rows from the independent column
y=y[:-forecast_out]

    #splitting the data into training and testing data
X_train, X_test, y_train, y_test = train_test_split(x, y, test_size=0.2) #20% testing

    # Linear Regression model
clflr=LinearRegression()
clflr.fit(X_train, y_train)

    # KNN Regression
clfknn = KNeighborsRegressor(n_neighbors=2)
clfknn.fit(X_train, y_train)

    # Bayesian Regression
clfbay = BayesianRidge(compute_score=True)
clfbay.fit(X_train, y_train)

    # SVM
clfsvm = svm.SVR(kernel='rbf')
clfsvm.fit(X_train, y_train)

    # Lasso Regression
clflas = Lasso()
clflas.fit(X_train, y_train)

    # create an array equal to the last 'n' number of days
x_forecast = np.array(df.drop(['Prediction'],1))[-forecast_out:]
model_pred = clfsvm.predict(x_forecast)

LR_pred = clflr.predict(X_test)
KNN_pred = clfknn.predict(X_test)
BAY_pred = clfbay.predict(X_test)
SVM_pred = clfsvm.predict(X_test)
LAS_pred = clflas.predict(X_test)

print("confidencereg: ", clflr.score(y_test, LR_pred))
print("confidencelas: ", clflas.score(y_test, LAS_pred))
print("confidencesvm: ", clfsvm.score(y_test, SVM_pred))
print("confidencebay: ", clfbay.score(y_test, BAY_pred))
print("confidenceknn: ", clfknn.score(y_test, KNN_pred))

print("Test LR MAE: ",metrics.mean_absolute_error(y_test,LR_pred))
print("Test LAS MAE: ",metrics.mean_absolute_error(y_test,LAS_pred))
print("Test SVM MAE: ",metrics.mean_absolute_error(y_test,SVM_pred))
print("Test BAY MAE: ",metrics.mean_absolute_error(y_test,BAY_pred))
print("Test KNN MAE: ",metrics.mean_absolute_error(y_test,KNN_pred))

    # create a list with the future 'n' number of days
dates = []
current = dt.date.today()
for i in range(11):
  current += dt.timedelta(days=1)
  dates.append(current)
print(dates)

fig = go.Figure()
fig.add_trace(
  go.Scatter(
    x=dates,
    y=model_pred.flatten(),
    mode='lines+markers',
    name='data'))
fig.update_layout(
  title="Predicted Close Price of next " + str(10) + " days",
  xaxis_title="Date",
  yaxis_title="Closed Price"
)