# Assigning Linear Regression to variable
model_sklearn = LinearRegression()


# Create a test set and training set
X_train, X_test, y_train, y_test = train_test_split(features_DSH.to_numpy().reshape(-1,1),
                                                    Target_DSH.to_numpy(), test_size= 0.10,
                                                    random_state= 30)

# Fitting the model to the training set
model_sklearn.fit(X_train, y_train)

# Predictions
y_pred = model_sklearn.predict(X_test)
printString = 'Residual as the Difference between Test Scores & Predicted Scores:'
printString_ = 'Residual using (r = y - X*beta ):'
printStr = 'Parameters (Slope/Intercept):'
parameters = np.array([model_sklearn.coef_[0], model_sklearn.intercept_])
# Output Dataframes
printPTS = pd.DataFrame.from_dict({'Predicted Test Scores': y_pred})
printParameters = pd.DataFrame.from_dict({printStr: parameters})
printR1 = pd.DataFrame.from_dict({printString: (y_test - y_pred)})
printR2 = pd.DataFrame.from_dict({printString_: finding_residuals(X_test, y_test, model_sklearn.intercept_, model_sklearn.coef_[0]).flatten()})
printRMSE = pd.DataFrame.from_dict({'RMSE:': [np.sqrt(metrics.mean_squared_error(y_test, y_pred))]})

# Outputting to file
count = 1
outputDataframeList = [printPTS, printParameters, printR1, printR2, printRMSE]
for outputDataframe in outputDataframeList:
    filepath = '/Users/Chris/Documents/Programming_Practice/gistHub/slrDFTTS_' + str(count) + '.csv'
    with open(filepath, 'w') as f:
        outputDataframe.to_csv(path_or_buf= f)
    count += 1