/
DataPreperation random.py
178 lines (150 loc) · 7 KB
/
DataPreperation random.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
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
import matplotlib.pyplot as plt
import numpy as np
import pandas as pd
from sklearn.impute import KNNImputer
from sklearn.ensemble import IsolationForest, RandomForestClassifier
from sklearn.metrics import accuracy_score
from ReliefF import ReliefF
from mlxtend.feature_selection import SequentialFeatureSelector
from sklearn.svm import SVC
from sklearn.feature_selection import SelectKBest
from sklearn.feature_selection import mutual_info_classif
from sklearn.neighbors import (NeighborhoodComponentsAnalysis,KNeighborsClassifier)
def Getwinner(res):
winner = []
for val in res:
if val > 0:
winner.append(1)
else:
winner.append(-1)
return pd.DataFrame(data=winner)
def AddColumns(data):
# Add random
data["random"] = pd.DataFrame(data=np.random.randint(-1, 2, size=(len(data), 1)))
#Add The number of goals the home group leads
res = data["fthg"] - data["ftag"]
data["The number of goals the home group leads"] = res
#Add winner
data["winner"] = Getwinner(res)
#data.to_csv("temp.csv")
def transform_to_numeric(data):
header = list(data.head(0))
header.remove("random")
header.remove("winner")
data = data.drop(columns=header)
return data
#split the data
def split_data(data):
train_data = data.head(int(60*len(data)/100))
valid_data = data.tail(int(40*len(data)/100)).head(int(20*len(data)/100))
test_data = data.tail(int(20*len(data)/100))
return train_data, valid_data, test_data
#Data Cleansing
def remove_outlier(data_X, data_Y):
# combine X and Y to consider both when detecting outliers
data = pd.concat([data_X, data_Y], axis=1)
clf = IsolationForest(random_state=0, contamination=0.08)
outlier_prediction = clf.fit_predict(data)
index_to_drop_list = []
for i in range(len(outlier_prediction)):
if outlier_prediction[i] == -1:
index_to_drop_list.append(i)
return data_X.drop(index_to_drop_list), data_Y.drop(index_to_drop_list)
def transform_season_to_one_hot(data):
one_hot = pd.get_dummies(data['season'], prefix='season')
data = data.drop(columns=['season'])
data = pd.concat([data, one_hot], axis=1)
return data
def min_max_scale(column, min, max):
if max - min == 0:
return column
return column.apply(lambda x: x / (max - min) )
def scale_data(train, valid, test):
min_max_features = list(train.head(0))
print(min_max_features)
for f in min_max_features:
max, min = train[f].max(), train[f].min()
train[f], valid[f], test[f] = min_max_scale(train[f],min, max), min_max_scale(valid[f],min, max), min_max_scale(test[f],min, max)
return train, valid, test
def plot_feature_with_label(data, label):
feature = [item for item in data.head(0)]
num = [30,31,32,33,34,35,36,37]
for i in num:
plt.figure()
plt.hist2d(data[feature[i]], label,bins=(50, 50), cmap=plt.cm.BuPu)
plt.ylabel('Label')
plt.xlabel(feature[i])
plt.show(block=True)
def check_performence(train_X, train_Y, valid_X, valid_Y):
clf = KNeighborsClassifier(n_neighbors=5)
clf.fit(train_X, train_Y)
res = clf.predict(valid_X)
print("accuracy on validation set: {}".format(accuracy_score(valid_Y, res)))
def filter_with_relief(train_x, valid_x, test_x, train_y):
temp_header = [item for item in train_x.head(0)]
features_to_drop = []
features_to_keep = 30
data, target = train_x.to_numpy(), train_y.to_numpy()
fs = ReliefF(n_neighbors=5, n_features_to_keep=features_to_keep)
fs.fit(data, target)
for index in fs.top_features[features_to_keep:]:
features_to_drop += [temp_header[index]]
return train_x.drop(features_to_drop, axis=1), valid_x.drop(features_to_drop, axis=1),\
test_x.drop(features_to_drop, axis=1)
def filter_with_sfs(train_X, valid_X, test_X, train_Y, i):
features = {item for item in train_X.head(0)}
fs = SequentialFeatureSelector(RandomForestClassifier(n_estimators=30, random_state=0),
k_features=i,
forward=True,
verbose=0,
scoring='accuracy',
cv=4)
fs.fit(train_X, train_Y)
selected_features = set(fs.k_feature_names_)
features_to_drop = list(features - selected_features)
return train_X.drop(features_to_drop, axis=1), valid_X.drop(features_to_drop, axis=1), \
test_X.drop(features_to_drop, axis=1)
def feature_selection(train_X, valid_X, test_X, train_Y, i):
# remove Linear dependencies feature
#train_X, valid_X, test_X = remove_linear_dependencies(train_X), remove_linear_dependencies(valid_X),\
# remove_linear_dependencies(test_X)
# using filter method
train_X, valid_X, test_X = filter_with_relief(train_X, valid_X, test_X, train_Y)
# using wrapper method
train_X, valid_X, test_X = filter_with_sfs(train_X, valid_X, test_X, train_Y, i)
return train_X, valid_X, test_X
def main():
raw_data = pd.read_csv("final data.csv", header=0)
# Add random and winner (as label)
AddColumns(raw_data)
numeric_data = transform_to_numeric(raw_data)
#numeric_data = transform_season_to_one_hot(numeric_data)
#numeric_data.to_csv("temp.csv")
train, valid, test = split_data(numeric_data)
train_Y, valid_Y, test_Y = train.winner, valid.winner, test.winner
#train_X, valid_X, test_X = train.drop(columns=['winner']), valid.drop(columns=['winner']), test.drop(columns=['winner'])
#check perfomans before startting preparation
################
#print("performence before:")
#check_performence(train_X, train_Y, valid_X, valid_Y)
##############
# check perfomans without the winning columns
train_X, valid_X, test_X = train.drop(columns=['winner']
), valid.drop(columns=['winner']
), test.drop(columns=['winner'])
print("performence before:")
check_performence(train_X, train_Y, valid_X, valid_Y)
train_X, train_Y = remove_outlier(train_X, train_Y)
print("performence after outlier dection:")
check_performence(train_X, train_Y, valid_X, valid_Y)
train_X, valid_X, test_X = scale_data(train_X, valid_X, test_X)
print("performence after scaling:")
check_performence(train_X, train_Y, valid_X, valid_Y)
t_train_X, t_valid_X, t_test_X = feature_selection(train_X, valid_X, test_X, train_Y, 30)
print("performence after " +str(i)+ " feature selection:")
check_performence(t_train_X, train_Y, t_valid_X, valid_Y)
final_train = pd.concat([train_Y, train_X], axis=1)
final_train.to_csv("processedTrainData.csv", index=False)
pd.DataFrame(data=train_X).to_csv("temp.csv")
if __name__ == "__main__":
main()