-
Notifications
You must be signed in to change notification settings - Fork 0
/
Code.py
128 lines (105 loc) · 2.91 KB
/
Code.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
sns.set(style='darkgrid')
from scipy.stats import pearsonr
from sklearn import metrics
from sklearn import preprocessing
from sklearn.linear_model import LinearRegression
from keras.utils import to_categorical
data = pd.read_csv("hour.csv").values
x = data[:,0]
y = data[:,11]
#-----------------Visualization---------------------#
y1 = data[:,1].reshape(17379,1)
y2 = data[:,2].reshape(17379,1)
y3 = data[:,3].reshape(17379,1)
y4 = data[:,4].reshape(17379,1)
y5 = data[:,5].reshape(17379,1)
y6 = data[:,6].reshape(17379,1)
y7 = data[:,7].reshape(17379,1)
y8 = data[:,8].reshape(17379,1)
y9 = data[:,9].reshape(17379,1)
y10 = data[:,10].reshape(17379,1)
y11 = data[:,11].reshape(17379,1)
y12 = data[:,12].reshape(17379,1)
y13 = data[:,13].reshape(17379,1)
y14 = data[:,14].reshape(17379,1)
y15 = data[:,15].reshape(17379,1)
y16 = data[:,16].reshape(17379,1)
# plt.plot(x,y2,x,y3)
# plt.show()
# plt.close()
# plt.plot(x,y4)
# plt.show()
# plt.close()
# plt.plot(x,y9)
# plt.show()
# plt.close()
# plt.plot(x,y6,x,y7)
# plt.show()
# plt.close()
# plt.plot(x,y10,x,y11)
# plt.show()
# plt.close()
###----------------------------------------------------###
##-------- Studying correlation between variables--------##
# print(pearsonr(y2,y3))
# # -0.0107
# print(pearsonr(y6,y7))
# #-0.102
# print(pearsonr(y7,y8))
# # 0.035
# print(pearsonr(y8,y9))
# #0.04
# print(pearsonr(y9,y10))
# # -0.1
# print(pearsonr(y10,y11))
# 0.98 : Highly correlated
# print(pearsonr(y10,y12))
# # -0.06
# print(pearsonr(y16,y15))
# casual, registered and total all highly correlated with each other
# Possible relation: Total = Linear combination of casual and registered.
# Model Selection and Training
y2 = y2-1
y4 = y4-1
# Converting categorical data into one-hot representation
y2 = to_categorical(y2)#,num_classes=4)
y4 = to_categorical(y4)#,num_classes=12)
y5 = to_categorical(y5)#,num_classes=24)
y7 = to_categorical(y7)#,num_classes=7)
y9 = to_categorical(y9)#,num_classes=4)
#print(y2.shape)
X1 = np.concatenate((y2,y3,y4,y5,y6,y7,y8,y9,y11,y12,y13),axis=1)
y = y16#data[:,16]
reg = LinearRegression()
reg.fit(X1,y)
print(reg.coef_)
r_square = reg.score(X1,y)
#print(X2.shape[1])
n = X1.shape[0]
p = X1.shape[1]
adjusted_r_square = 1-(1 - r_square)*(n-1)/(n-p-1)
print(r_square)
print(adjusted_r_square)
###-----------------------------------------------###
# Time series approach
from statsmodels.tsa.arima_model import ARIMA
#from statsmodels.tsa.statespace import SARIMAX
model = ARIMA(y16,order=(4,0,2))
model_fit = model.fit(disp=0)
print(model_fit.summary())
# residuals = pd.DataFrame(model_fit.resid)
# residuals.plot()
# plt.show()
# residuals.plot(kind='kde')
# plt.show()
# print(residuals.describe())
# mod = SARIMAX(y16,order=(1,0,1),seasonal_order=(1,1,1,4))
# results = mod.fit()
# print(results.summary())
#plt.plot(x[0:1808],y16[0:1808])
#plt.show()
#plt.close()