-
Notifications
You must be signed in to change notification settings - Fork 0
/
HW3_VMelland.py
212 lines (165 loc) · 9.31 KB
/
HW3_VMelland.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
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
import arff, numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.naive_bayes import GaussianNB
import matplotlib.pyplot as plt
from sklearn.metrics import confusion_matrix
from sklearn.metrics import classification_report
import seaborn as sns
import math
import random
from sklearn import tree
from sklearn.tree import DecisionTreeClassifier
# Age: Any ages in years when a women during pregnant.
# SystolicBP: Upper value of Blood Pressure in mmHg, another significant attribute during pregnancy.
# DiastolicBP: Lower value of Blood Pressure in mmHg, another significant attribute during pregnancy.
# BS: Blood glucose levels is in terms of a molar concentration, mmol/L.
# HeartRate: A normal resting heart rate in beats per minute.
# Risk Level: Predicted Risk Intensity Level during pregnancy considering the previous attribute.
df = pd.read_csv("Maternal Health Risk Data Set.csv")
print(df)
# percentile list
perc =[.20, .40, .60, .80]
# list of dtypes to include
include =['object', 'float', 'int']
# calling describe method
desc = df.describe(percentiles = perc, include = include)
# display
print(desc)
#print(df)
X_train, X_test, y_train, y_test = train_test_split(df[['Age', 'SystolicBP', 'DiastolicBP', 'BS', 'BodyTemp', 'HeartRate']],
df['RiskLevel'], random_state=0)
NB = GaussianNB()
NB.fit(X_train, y_train)
y_predict = NB.predict(X_test)
print("Accuracy Test Set NB: {:.2f}".format(NB.score(X_test, y_test)))
print("Accuracy Train Set NB: {:.2f}".format(NB.score(X_train, y_train)))
print('Classification report \n', classification_report(y_test, y_predict))
cm = confusion_matrix(y_test, y_predict)
print('Confusion matrix\n\n', cm)
print('\nTrue Positives(TP) = ', cm[0,0])
print('\nTrue Negatives(TN) = ', cm[2,1] + cm[2,2] + cm[1,2] + cm[1,1])
print('\nFalse Positives(FP) = ', cm[1,0] + cm[2,0])
print('\nFalse Negatives(FN) = ', cm[0,1] + cm[0,2])
##################################################################################
#Trying to redo the Naive Bayes but from scratch without packages
#prior = P('high risk')
#P(high risk) = how many high risk appears / total observations
#P(risk levels) for low high and mid risk
prior = df.groupby(by = "RiskLevel").size().div(len(df))
#prob values grouped by risk level
#print(prior)
#next calculate likelihood for each of the features in the dataset
#P(Age|risk level)
#P(SystolicBP|risk level)
#P(DiastolicBP| risk level)
#P(BS | risk level)
#P(HeartRate | risk level)
likelihood = {}
likelihood['BodyTemp'] = df.groupby(['RiskLevel', 'BodyTemp']).size().div(len(df)).div(prior)
likelihood['HeartRate'] = df.groupby(['RiskLevel', 'HeartRate']).size().div(len(df)).div(prior)
likelihood['BS'] = df.groupby(['RiskLevel', 'BS']).size().div(len(df)).div(prior)
likelihood['DiastolicBP'] = df.groupby(['RiskLevel', 'DiastolicBP']).size().div(len(df)).div(prior)
likelihood['SystolicBP'] = df.groupby(['RiskLevel', 'SystolicBP']).size().div(len(df)).div(prior)
likelihood['Age'] = df.groupby(['RiskLevel', 'Age']).size().div(len(df)).div(prior)
#print(likelihood)
#sorting everything by values lowest to highest
example = df.sort_values(by=['Age','SystolicBP', 'DiastolicBP', 'BS', 'BodyTemp', 'HeartRate'])
#print(example)
#sorting the data for ages <= 25 and more than 25
less_than25 = example[example['Age'] <= 25]
#print(less_than25)
more_than25 = example[example['Age'] > 25]
#print(more_than25)
#prob of low high and midrisk based on data
# p_highrisk = likelihood['Age']['']
# p_lowrisk =
# p_midrisk =
prior_lt25 = less_than25.groupby(by = "RiskLevel").size().div(len(less_than25))
#prob values grouped by risk level
print("RISK LEVEL PROBABILITY FOR UNDER 25 GROUP\n")
print(prior_lt25)
prior_mt25 = more_than25.groupby(by = "RiskLevel").size().div(len(more_than25))
#prob values grouped by risk level
print("RISK LEVEL PROBABILITY FOR OVER 25 GROUP\n")
print(prior_mt25)
print("As you can see, there is a higher probability that the over 25 group has high or mid risk people compared to the less than 25 group")
likelihood2 = {}
likelihood2['BodyTemp'] = less_than25.groupby(['RiskLevel', 'BodyTemp']).size().div(len(less_than25)).div(prior_lt25)
likelihood2['HeartRate'] = less_than25.groupby(['RiskLevel', 'HeartRate']).size().div(len(less_than25)).div(prior_lt25)
likelihood2['BS'] = less_than25.groupby(['RiskLevel', 'BS']).size().div(len(less_than25)).div(prior_lt25)
likelihood2['DiastolicBP'] = less_than25.groupby(['RiskLevel', 'DiastolicBP']).size().div(len(less_than25)).div(prior_lt25)
likelihood2['SystolicBP'] = less_than25.groupby(['RiskLevel', 'SystolicBP']).size().div(len(less_than25)).div(prior_lt25)
likelihood2['Age'] = less_than25.groupby(['RiskLevel', 'Age']).size().div(len(less_than25)).div(prior_lt25)
likelihood3 = {}
likelihood3['BodyTemp'] = more_than25.groupby(['RiskLevel', 'BodyTemp']).size().div(len(more_than25)).div(prior_mt25)
likelihood3['HeartRate'] = more_than25.groupby(['RiskLevel', 'HeartRate']).size().div(len(more_than25)).div(prior_mt25)
likelihood3['BS'] = more_than25.groupby(['RiskLevel', 'BS']).size().div(len(more_than25)).div(prior_mt25)
likelihood3['DiastolicBP'] = more_than25.groupby(['RiskLevel', 'DiastolicBP']).size().div(len(more_than25)).div(prior_mt25)
likelihood3['SystolicBP'] = more_than25.groupby(['RiskLevel', 'SystolicBP']).size().div(len(more_than25)).div(prior_mt25)
likelihood3['Age'] = more_than25.groupby(['RiskLevel', 'Age']).size().div(len(more_than25)).div(prior_mt25)
#print(likelihood3)
#the likelihoods can be printed if you want, Likelihood1 is before sorting everything
#like2 is for <=25 and like3 is for >25
###############################################################
#Trying to retest the NB to see if using the split up data is better than before
X_train, X_test, y_train, y_test = train_test_split(less_than25[['Age', 'SystolicBP', 'DiastolicBP', 'BS', 'BodyTemp', 'HeartRate']],
less_than25['RiskLevel'], random_state=0)
NB = GaussianNB()
NB.fit(X_train, y_train)
y_predict = NB.predict(X_test)
print("Accuracy Test Set NB: {:.2f}".format(NB.score(X_test, y_test)))
print("Accuracy Train Set NB: {:.2f}".format(NB.score(X_train, y_train)))
print('Classification report for Less-Than 25 Group \n', classification_report(y_test, y_predict))
X_train, X_test, y_train, y_test = train_test_split(more_than25[['Age', 'SystolicBP', 'DiastolicBP', 'BS', 'BodyTemp', 'HeartRate']],
more_than25['RiskLevel'], random_state=0)
NB = GaussianNB()
NB.fit(X_train, y_train)
y_predict = NB.predict(X_test)
print("Accuracy Test Set NB: {:.2f}".format(NB.score(X_test, y_test)))
print("Accuracy Train Set NB: {:.2f}".format(NB.score(X_train, y_train)))
print('Classification report for More-Than 25 Group \n', classification_report(y_test, y_predict))
print("If I split up the data into two separate data tables, one for people ages less than 25 and the other for ages more than 25, it gives a higher accuracy of classifying the risk levels of the groups.")
###########################################################################
#Create a Decision Tree for the data
#Going to rename the high mid and low risk values to:
# Low Risk = 0
# Mid Risk = 1
# High Risk = 2
d = {'low risk':0, 'mid risk':1, 'high risk':2}
df['RiskLevel'] = df['RiskLevel'].map(d)
print(df)
#Then separate the feature and target columns
features = ['Age', 'SystolicBP', 'DiastolicBP', 'BS', 'BodyTemp', 'HeartRate']
x = df[features]
y = df['RiskLevel']
# print(x)
# print(y)
dtree = DecisionTreeClassifier(max_depth = 5, min_samples_leaf=5)
dtree = dtree.fit(x,y)
axe = plt.subplots(figsize=(20,10))
tree.plot_tree(dtree, feature_names = features)
plt.savefig('out.pdf')
#Using train test split to make the decision tree, splitting up the data into less than 25
X_train, X_test, y_train, y_test = train_test_split(less_than25[['Age', 'SystolicBP', 'DiastolicBP', 'BS', 'BodyTemp', 'HeartRate']],
less_than25['RiskLevel'], random_state=0)
tree_lt25 = dtree.fit(X_train, y_train)
axe = plt.subplots(figsize=(20,10))
tree.plot_tree(tree_lt25, feature_names = features)
plt.savefig('outlt25.pdf')
#VALUE in the results can be a low, mid, or high risk.
#value = [low risk, mid risk, high risk]
#using the train test split to make dec. tree using the split data for people more than 25
X_train, X_test, y_train, y_test = train_test_split(more_than25[['Age', 'SystolicBP', 'DiastolicBP', 'BS', 'BodyTemp', 'HeartRate']],
more_than25['RiskLevel'], random_state=0)
tree_MT25 = dtree.fit(X_train, y_train)
axe = plt.subplots(figsize=(20,10))
tree.plot_tree(tree_MT25, feature_names = features)
plt.savefig('outMT25.pdf')
X_train, X_test, y_train, y_test = train_test_split(df[['Age', 'SystolicBP', 'DiastolicBP', 'BS', 'BodyTemp', 'HeartRate']],
df['RiskLevel'], random_state=0)
tree_ALLtts = dtree.fit(X_train, y_train)
axe = plt.subplots(figsize=(20,10))
tree.plot_tree(tree_ALLtts, feature_names = features)
plt.savefig('outALL.pdf')
# print(df)