Create new file
Browse files- diabetes-model-pl.py +36 -0
diabetes-model-pl.py
ADDED
@@ -0,0 +1,36 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Importing essential libraries
|
2 |
+
import numpy as np
|
3 |
+
import pandas as pd
|
4 |
+
import pickle
|
5 |
+
|
6 |
+
# Loading the dataset
|
7 |
+
df = pd.read_csv('plashkar/diabetes-db')
|
8 |
+
|
9 |
+
# Renaming DiabetesPedigreeFunction as DPF
|
10 |
+
df = df.rename(columns={'DiabetesPedigreeFunction':'DPF'})
|
11 |
+
|
12 |
+
# Replacing the 0 values from ['Glucose','BloodPressure','SkinThickness','Insulin','BMI'] by NaN
|
13 |
+
df_copy = df.copy(deep=True)
|
14 |
+
df_copy[['Glucose','BloodPressure','SkinThickness','Insulin','BMI']] = df_copy[['Glucose','BloodPressure','SkinThickness','Insulin','BMI']].replace(0,np.NaN)
|
15 |
+
|
16 |
+
# Replacing NaN value by mean, median depending upon distribution
|
17 |
+
df_copy['Glucose'].fillna(df_copy['Glucose'].mean(), inplace=True)
|
18 |
+
df_copy['BloodPressure'].fillna(df_copy['BloodPressure'].mean(), inplace=True)
|
19 |
+
df_copy['SkinThickness'].fillna(df_copy['SkinThickness'].median(), inplace=True)
|
20 |
+
df_copy['Insulin'].fillna(df_copy['Insulin'].median(), inplace=True)
|
21 |
+
df_copy['BMI'].fillna(df_copy['BMI'].median(), inplace=True)
|
22 |
+
|
23 |
+
# Model Building
|
24 |
+
from sklearn.model_selection import train_test_split
|
25 |
+
X = df.drop(columns='Outcome')
|
26 |
+
y = df['Outcome']
|
27 |
+
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.20, random_state=0)
|
28 |
+
|
29 |
+
# Creating Random Forest Model
|
30 |
+
from sklearn.ensemble import RandomForestClassifier
|
31 |
+
classifier = RandomForestClassifier(n_estimators=20)
|
32 |
+
classifier.fit(X_train, y_train)
|
33 |
+
|
34 |
+
# Creating a pickle file for the classifier
|
35 |
+
filename = 'diabetes-prediction-rfc-model.pkl'
|
36 |
+
pickle.dump(classifier, open(filename, 'wb'))
|