-
Notifications
You must be signed in to change notification settings - Fork 0
/
Diabetes_Disease_Prediction.py
144 lines (107 loc) · 5.94 KB
/
Diabetes_Disease_Prediction.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
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.model_selection import train_test_split
from sklearn.metrics import accuracy_score,confusion_matrix
from flask import Flask, request, jsonify
dataframe=pd.read_csv("diabetes.csv")
# dataframe.head(10)
# dataframe.info()
# dataframe.describe()
# dataframe.isnull().sum()
# Spliting the data
Diabetes_X= dataframe.iloc[:,:-1].values
Diabetes_Y= dataframe.iloc[:,-1].values
x_train,x_test,y_train,y_test=train_test_split(Diabetes_X,Diabetes_Y,test_size=0.2,random_state=23)
from sklearn.ensemble import RandomForestClassifier
rfc=RandomForestClassifier(n_estimators=500,criterion='entropy',max_depth=8,min_samples_split=5)
randomFC=RandomForestClassifier()
randomFC=rfc.fit(x_train,y_train)
f1_pred=randomFC.predict(x_test)
# print("Predicted",f1_pred)
f1_score=accuracy_score(y_test,f1_pred)*100
# print("Accuracy_RFC",f1_score)
# -----------------------------------------------------------------------------------------
# Not using neural network algo because it's accuracy is lower than the RFC in predicting diabetes
x_train,x_test,y_train,y_test=train_test_split(Diabetes_X,Diabetes_Y,test_size=0.2,random_state=42)
from sklearn.neural_network import MLPClassifier
neuralNet = MLPClassifier()
neuralNet=rfc.fit(x_train, y_train)
prediction5 = neuralNet.predict(x_test)
accuracy_nn = accuracy_score(y_test, prediction5)
# print("Accuracy NeuralNET",accuracy_nn)
# ---------------------------------------------------------------------------------------------------
def diabetes_prediction():
importance_data_dict = {}
advice_text=""
if request.is_json:
form_data = request.get_json() # Get JSON data from the request
# user_input = form_data['user_input'] # Extract user input from form data
user_input_list = [float(value) for value in form_data.values()]
# --------------------------------------------------------------------------
# Perform any processing with user_input here
# Now predicting the Heart Disease on the bases of usr input
input_data = np.array(user_input_list).reshape(1, -1)
pre1 = rfc.predict(input_data)
if pre1 == 1:
response_text="The patient seems to be have Diabetes."
advice_text=" We recommend monitoring and controlling these factors to manage your risk of Diabetes"
feature_importances = rfc.feature_importances_
importance_df = pd.DataFrame({'Feature': dataframe.columns[:-1], 'Importance': feature_importances})
importance_df = importance_df.sort_values(by='Importance', ascending=False)
importance_data_dict = importance_df.to_dict()
# 5. Provide Recommendations
else:
response_text="The patient seems to be Normal..."
# For now, just return the received data
return jsonify({
# "received_data": user_input,
# "pridicted_result": response_text
"pridicted_result" :response_text,
"importance_data": importance_data_dict,
"advice_text":advice_text
})
else:
return jsonify({'error':'invalid request format, expected JSON'})
# # Takig usr input of Diabetes causing Parameters through terminal
# user_input = []
# for i in range(len(dataframe.columns) - 1): # Exclude the last column (Outcome)
# column_name = dataframe.columns[i] # Get the name of the column
# value = float(input("Enter no of {}: ".format(column_name)))
# user_input.append(value)
# # Now predicting the Diabetes on the bases of usr input
# input_data = np.array(user_input).reshape(1, -1)
# prediction = randomFC.predict(input_data)
# # if prediction == 1:
# # print("The model predicts that the user may have diabetes.")
# # else:
# # print("The model predicts that the user may not have diabetes.")
# # Getting all the major parameter that are highly reponsible for causing the diabetes
# feature_importances = randomFC.feature_importances_
# importance_df = pd.DataFrame({'Feature': dataframe.columns[:-1], 'Importance': feature_importances})
# importance_df = importance_df.sort_values(by='Importance', ascending=False)
# print(importance_df)
# if prediction == 1:
# feature_contributions = randomFC.predict_proba(input_data)[0] # Get the probabilities for each class
# # The first element of feature_contributions corresponds to the probability of class 0 (no diabetes)
# # The second element corresponds to the probability of class 1 (diabetes)
# # We can consider the difference between the probabilities as the contribution of each feature to the predicted risk of diabetes
# # The higher the difference, the more influential the feature is in predicting diabetes
# feature_contribution_diff = feature_contributions[1] - feature_contributions[0]
# # 4. Identify Highly Contributing Features
# # We can consider features with higher contribution differences as highly contributing features
# highly_contributing_features = []
# for i, feature_name in enumerate(dataframe.columns[:-1]): # Exclude the last column (Outcome)
# if feature_contribution_diff > 0.127385: # Adjust the threshold as needed
# highly_contributing_features.append(feature_name)
# # 5. Provide Recommendations
# if len(highly_contributing_features) > 0:
# print("Based on the analysis of your input data, the following features are highly contributing to the risk of diabetes:")
# for feature_name in highly_contributing_features:
# print("- {}".format(feature_name))
# print("We recommend monitoring and controlling these factors to manage your risk of diabetes.")
# else:
# print("No specific features were identified as highly contributing to the risk of diabetes based on your input data.")
# else:
# print("Based on the analysis of your input data, it seems that you have a low risk of diabetes.")