Linear Regression implementation using python

Linear Regression Implementation:

In the previous post we saw about the theoretical explanation about the linear regression. Now let's see it's implementation using python. 

If you haven't read my post about linear regression, then click here

Data set used : - Student Grades

 

import numpy as np

import pandas as pd

import sklearn

from sklearn import linear_model

from sklearn.utils import shuffle

data = pd.read_csv("student-mat.csv",sep=';')

data = data[['G1','G2','G3','studytime','failures','absences']]

print(data.head())  ''' head() prints the first tuples of the data default value is 5 '''

predict = "G3"

X = np.array(data.drop([predict],1)) #features

y = np.array(data[predict]) #labels

''' this is gonna split the data into test and train data with 10% as the test data size '''

x_train, x_test, y_train, y_test = sklearn.model_selection.train_test_split(X,y,test_size=0.1)

model = linear_model.LinearRegression() #create the model

''' Train the model'''

model.fit(x_train,y_train) 

accuracy = model.score(x_test,y_test) ''' This gives the accuracy of the model to predict the values '''

print("Accuraccy {}".format(accuracy))

print("Coefficient :",model.coef_) ''' Gives the slope '''

print("Intercept :",model.intercept_)

predictions = model.predict(x_test) 

print('Predictions')

for x in range(len(predictions)):

    print(predictions[x],x_test[x],y_test[x])

As you can see form the below outputs, when we executed the program for the second time the accuracy level dropped form 89% to 81%. So, we need to save our model with maximum accuracy. 

How to save the model with maximum accuracy?

Output:-

  G1  G2  G3  studytime  failures  absences
0 5   6   6      2         0         6
1 5   5   6      2         0         4
3 15  14  15     3         0         2
2 7   8  10      2         3        10
4 6  10  10      2         0         4

Accuraccy 0.8982728958031367

Coefficient : [ 0.16172694  0.98027571 -0.20876813 -0.31050997  0.03495033]
Intercept : -1.5754139422756488

Predictions
7.8160282292435745 [10  8  2  0 10] 8
10.638979646447401 [11 11  2  0  2] 10
6.371563440971993 [8 7 1 0 0] 6
10.465161852407348 [11 11  3  0  3] 11
12.597274977578584 [14 13  4  0  0] 14
14.042128123637497 [13 14  1  0  0] 13
9.505143830469267 [ 7 10  2  1 25] 11
10.339319107958076 [10 11  4  0 10] 11
13.928376231032752 [11 14  1  0  6] 14
12.081576703155246 [13 12  1  0  0] 12
7.400584945046997 [8 9 1 3 0] 10
11.27294199217762 [ 8 12  1  0  0] 12
9.196382589739558 [ 9 10  3  0  4] 10
5.002910063736628 [ 6  6  2  1 13] 8
7.07503809284365 [ 8  8  4  0 10] 8
6.302596642825093 [8 7 2 0 4] 6
12.922984951320702 [13 13  2  0  2] 11
8.332114861454244 [8 9 1 0 0] 8
12.733542158634021 [14 12  2  0 20] 13
16.187265966435127 [15 16  2  0  2] 16
11.892133910468566 [14 11  1  0 18] 13
5.792300246767593 [7 7 3 0 0] 8
3.8568643957562667 [5 5 2 0 4] 6
18.239643665184136 [16 18  2  0  0] 18
-1.9070772737097068 [5 0 1 3 0] 0
13.834293863414928 [13 14  3  0  6] 14
11.632199268943827 [13 12  2  1  2] 12
19.961843377607057 [18 19  1  0  6] 19
15.206990256194004 [15 15  2  0  2] 14
8.263148063307344 [8 9 2 0 4] 10
11.98975042488849 [12 12  1  0  2] 14
10.239834861974414 [12 10  2  0 14] 11
8.446800619840857 [10  9  2  0  0] 9
15.125932606202575 [16 15  4  0  7] 17
7.550301840769091 [7 9 2 2 6] 8
3.521238854149594 [7 5 3 1 0] 0
11.121983678257424 [ 9 12  3  0  3] 11
-1.0696297249202846 [7 0 3 0 0] 0
8.42487500550287 [9 9 2 0 4] 10
5.760459064565051 [7 7 2 1 2] 7

Output 2:

  G1  G2  G3  studytime  failures  absences
0 5   6   6      2         0         6
2 7   8  10      2         3        10
1 5   5   6      2         0         4
3 15  14  15     3         0         2
4 6  10  10      2         0         4

Accuraccy 0.8130042352757954

Coefficient : [ 0.14958581  0.98020592 -0.23107325 -0.28833124  0.04045218]
Intercept : -1.3617056273559687

Predictions
12.195249504944599 [10 13  4  0  6] 13
7.0643128006959675 [8 8 3 0 2] 10
9.752064477957072 [10 10  2  1 14] 9
10.696984164357781 [10 11  2  0  6] 11
15.677131419596725 [16 15  2  0 10] 15
8.493859213317853 [10  9  2  0  0] 0
9.465294725293901 [ 9  9  1  0 22] 9
10.154516368964485 [ 9 11  3  0  2] 11
3.6863610639193425 [6 5 2 1 0] 0
-1.0078247421768198 [7 0 3 0 0] 0
-0.9263372968086295 [6 0 2 0 0] 0
11.896040401967113 [11 12  1  0  2] 11
15.2039281355093 [15 15  2  0  2] 14
15.688771277982474 [16 15  3  0 16] 15
15.041536320449486 [16 15  3  0  0] 15
8.000066737921825 [6 9 1 1 4] 8
9.24299188361126 [10 10  3  0  0] 9
7.001089365007859 [10  8  1  3  3] 7
6.928624546719691 [ 6  8  1  3 16] 8
5.537818209325654 [7 6 1 0 5] 7
9.820745913586379 [11 10  2  1 12] 10
15.2039281355093 [15 15  2  0  2] 16
13.774381722887414 [13 14  3  0  4] 14
10.396646402362787 [10 11  4  0 10] 11
5.335557285096596 [7 6 1 0 0] 0
9.554969503992005 [10 10  2  0  2] 10
5.268183807325057 [ 6  6  2  2 22] 4
14.01826098331542 [11 14  1  0  6] 14
4.136501047487263 [6 5 2 0 4] 6
9.94668534271124 [12 10  3  0 10] 12
10.995572699323075 [13 11  3  0  8] 11
13.768480763260206 [11 12  2  0 54] 11
13.280185095731193 [12 13  2  0 14] 12
12.126530576979667 [12 12  1  0  4] 13
16.40240129744207 [17 16  2  0  0] 17
17.908485621420148 [17 17  2  0 13] 17
10.1417103592256 [11 11  4  0  0] 11
9.405383698671075 [ 9 10  2  0  2] 9
8.545951256549415 [10  9  3  0  7] 9
6.476988869785805 [8 7 2 0 6] 9