LINEAR REGRESSOR

This is a module of numerical statistic prediction using linear regression by determinant matrix method with intercept, slope, correlation coefficient, and determination coefficient calculations

By Zaafirrahman

Installation

This library can be installed by the pip command, open your terminal and type in the following command...

pip install linearregressor

Functions of this library

First import the library using this command

import linearregressor as lr

and then proceed to call the functions

The parameters

x (train independent variable) - Dataframe of independent variable from train data

y (train dependent variable) - Series of dependent variable from train data

test (test independent variable) - Dataframe of independent variable from test data

Prediction

lr.predict(x,y,test)

This function can be used to predict dependent test data from given independent test data (test) by coefficient calculation from given dependent (y) and independent train data (x)

Example :

Given a train data with correlation between dependent and independent variable as shown below:

X1	X2	Y
10	7	23
2	3	7
4	2	15
6	4	17
8	6	23
7	5	22
4	3	10
6	3	14
7	4	20
6	3	19

Based on train data above, predict the dependent value from test data below:

X1	X2
9	6
5	4
7	5
8	6
5	6
4	2
9	5
10	8
6	7
7	3

Code:

# Create train data as pandas dataframe format
df = pd.DataFrame({'Y':[23,7,15,17,23,22,10,14,20,19],
                   'X1':[10,2,4,6,8,7,4,6,7,6],
                   'X2':[7,3,2,4,6,5,3,3,4,3]})
# For simplest way, you can just import your csv file using pd.read_csv() function

# Split between dependent and independent data
y = df['Y']
x = df.drop(['Y'],axis=1)

# Create test data as pandas dataframe format and predict the dependent value
df_test = pd.DataFrame({'X1':[9,5,7,8,5,4,9,10,6,7],
                        'X2':[6,4,5,6,6,2,5,8,7,3]})
lr.predict(x,y,df_test)

Output:

[23.540636042402866,
 14.508833922261505,
 19.024734982332188,
 21.04946996466434,
 13.575971731448758,
 12.950530035335726,
 24.00706713780924,
 25.098939929328644,
 15.600706713780912,
 19.957597173144933]

Intercept

lr.intercept(x,y)

This function can be used to find intercept value from dependent (y) and independent train data (x)

Example:

With the same train data, you can find the intercept value of the data

>>> lr.intercept(x,y)
3.918727915194365

As shown in the output, it can be concluded that if the value of the independent variable is constant, then the average value of the dependent variable is 3.918727915194365

Slope

lr.slope(x,y)

This function can be used to find slope value from dependent (y) and independent train data (x)

Example:

With the same train data, you can find the slope value of the data

>>> lr.slope(x,y)
[2.4911660777385274, -0.46643109540637395]

As shown in the output, it can be concluded that if the value of X1 increases by 1 unit, it will increase the value of the dependent variable by 2.4911660777385274 because the coefficient is marked (+) which means the effect is positive. Meanwhile, if the value of X2 increases by 1 unit, it will decrease the value of the dependent variable by 0.46643109540637395 because the coefficient is marked (-) which means the effect is negative.

Correlation Coefficient

lr.r(x,y)

This function can be used to find correlation coefficient from dependent (y) and independent train data (x)

Example:

With the same train data, you can find the correlation coefficient of the data

>>> lr.r(x,y)
0.9145723194701728

As shown in the output, it can be concluded that between the independent variable and the dependent variable, which has a very strong correlation because the correlation coefficient is very close to 1.

Determination Coefficient

lr.rsquare(x,y)

This function can be used to find determination coefficient from dependent (y) and independent train data (x)

Example:

With the same train data, you can find the determination coefficient of the data

>>> lr.rsquare(x,y)
0.8364425275410518

As shown in the output, it can be concluded that the value of the independent variable affects 0.8364425275410518 or 83.64% of the value of the dependent variable, while the remaining 16.36% is influenced by other variables.

Regression Information

lr.info(x,y)

This function can be used to easily show the information of your data such as constant (intercept), coefficient (slope), multiple r (correlation coefficient), and r squared (determination coefficient) together in just one function

Example:

With the same train data, you can show the whole regression information about your data

>>> lr.info(x,y)
Constant :  3.918727915194365
Coefficient_1 : 2.4911660777385274
Coefficient_2 : -0.46643109540637395
Multiple R :  0.9145723194701728
R Squared :  0.8364425275410518