Building Linear Regression model

Linear regression is arguably the simplest model that one can build. It is a model of choice if you know that the relationship between your dependent variable and independent ones is linear.

Getting ready

To execute this recipe, you will need pandas, NumPy, and Scikit. No other prerequisites are required.

How to do it…

Estimating the regression model with Scikit is extremely easy (the regression_linear.py file):

import sys
sys.path.append('..')

# the rest of the imports
import helper as hlp
import pandas as pd
import numpy as np
import sklearn.linear_model as lm

@hlp.timeit
def regression_linear(x,y):
    '''
        Estimate a linear regression
    '''
    # create the regressor object
    linear = lm.LinearRegression(fit_intercept=True,
        normalize=True, copy_X=True, n_jobs=-1)

    # estimate the model
    linear.fit(x,y)

    # return the object
    return linear

# the file name of the dataset
r_filename = '../../Data/Chapter6/power_plant_dataset_pc...