Scribd
Upload
7 views

ml2020 Pythonlab02

This document discusses linear regression and describes applying various linear regression methods to a random dataset. It introduces linear regression, imports necessary packages, downloads a house price dataset, explores the data, splits it into training and test sets, fits a linear regression model to the training set, and prints the intercept and coefficients. It then provides an exercise to generate a random dataset, apply different linear regression methods to it including scipy.Polyfit, stats.linregress, optimize.curve_fit, numpy.linalg.lstsq, statsmodels.OLS, and sklearn.LinearRegression, show the results on a graph and find the fastest

Uploaded by

VINAY U PAI
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
7 views

ml2020 Pythonlab02

This document discusses linear regression and describes applying various linear regression methods to a random dataset. It introduces linear regression, imports necessary packages, downloads a house price dataset, explores the data, splits it into training and test sets, fits a linear regression model to the training set, and prints the intercept and coefficients. It then provides an exercise to generate a random dataset, apply different linear regression methods to it including scipy.Polyfit, stats.linregress, optimize.curve_fit, numpy.linalg.lstsq, statsmodels.OLS, and sklearn.LinearRegression, show the results on a graph and find the fastest

Uploaded by

VINAY U PAI
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

Birla Institute of Technology and Science, Pilani

Department of computer science & information system


BITS F464 - Machine Learning
I Semester 2020-21
28-Aug-20 Lab Sheet-02 – Linear Regression

Introduction

In statistics, linear regression is a linear approach for modeling the relationship between a scalar dependent
variable y and one or more explanatory variables (or independent variables) denoted X. The case of one
explanatory variable is called simple linear regression. For more than one explanatory variable, the process
is called multiple linear regression.

Linear regression models are often fitted using the least squares approach, but they may also be fitted in
other ways, such as by minimizing the "lack of fit" in some other norm (as with least absolute deviations
regression), or by minimizing a penalized version of the least squares loss function as in ridge regression
(L2-norm penalty) and lasso (L1-norm penalty). Conversely, the least squares approach can be used to fit
models that are not linear models. Thus, although the terms "least squares" and "linear model" are closely
linked, they are not synonymous.

Import packages and download the US house price dataset from


“https://www.kaggle.com/farhankarim1/usa-house-prices”

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline

df = pd.read_csv("your dataset location")


df.head()

Check basic info on the data set


'info()' method to check the data types and number
df.info(verbose=True)

'describe()' method to get the statistical summary of the various features of the data set
df.describe(percentiles=[0.1,0.25,0.5,0.75,0.9])

'columns' method to get the names of the columns (features)


df.columns

Basic plotting and visualization on the data set


Pairplots using seaborn
sns.pairplot(df)
Distribution of price (the predicted quantity)
df['Price'].plot.hist(bins=25,figsize=(8,4))
df['Price'].plot.density()

Correlation matrix and heatmap

df.corr()
plt.figure(figsize=(10,7))
sns.heatmap(df.corr(),annot=True,linewidths=2)

Feature and variable sets

Make a list of data frame column names


l_column = list(df.columns) # Making a list out of column names
len_feature = len(l_column) # Length of column vector list
l_column

Put all the numerical features in X and Price in y, ignore Address which is string for linear regression

X = df[l_column[0:len_feature-2]]
y = df[l_column[len_feature-2]]
print("Feature set size:",X.shape)
print("Variable set size:",y.shape)

X.head()
y.head()

Test-train split
Import train_test_split function from scikit-learn
from sklearn.cross_validation import train_test_split

Create X and y train and test splits in one command using a split ratio and a random seed
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=123)

Check the size and shape of train/test splits (it should be in the ratio as per test_size parameter above)
print("Training feature set size:",X_train.shape)
print("Test feature set size:",X_test.shape)
print("Training variable set size:",y_train.shape)
print("Test variable set size:",y_test.shape)

Model fit and training


Import linear regression model estimator from scikit-learn and instantiate

from sklearn.linear_model import LinearRegression


from sklearn import metrics

lm = LinearRegression() # Creating a Linear Regression object 'lm'

Fit the model on to the instantiated object itself and Check the intercept and coefficients and put them in a
DataFrame
lm.fit(X_train,y_train) # Fit the linear model on to the 'lm' object itself i.e. no need to set this to another variable
print("The intercept term of the linear model:", lm.intercept_)
print("The coefficients of the linear model:", lm.coef_)

#idf = pd.DataFrame(data=idict,index=['Intercept'])
cdf = pd.DataFrame(data=lm.coef_, index=X_train.columns, columns=["Coefficients"])
#cdf=pd.concat([idf,cdf], axis=0)
cdf

Lab 02 Exercise (Submit the code in given time):


These are the various linear regression methods:
 Scipy.Polyfit
 Stats.linregress
 Optimize.curve_fit
 numpy.linalg.lstsq
 statsmodels.OLS
 Analytic solution using Moore-Penrose generalized inverse or simple multiplicative matrix inverse
 sklearn.linear_model.LinearRegression

Generate a random dataset of some points and apply all the above mentioned regression methods on this
dataset. Show the results on a graph. Also, find the most suitable method according to the execution time.

You might also like