ashna-rajput/sharing-github

Machine Learning models using Ensemble modelling

import pandas as pd import numpy as np

import seaborn as sns #grid search is used to find the optimal paramters for each model from sklearn.model_selection import GridSearchCV from sklearn import metrics #Importing KNN, Logistic and Rf classifier from sklearn library from sklearn.neighbors import KNeighborsClassifier #Voting Classifier is used to make the prediction by majority vote from sklearn.ensemble import VotingClassifier, RandomForestClassifier from sklearn.linear_model import LogisticRegression from sklearn.model_selection import train_test_split

import matplotlib.pyplot as plt

Reading the csv dataset

df = pd.read_csv('C:/Users/AWU/Desktop/CSV Files/diabetes.csv')

Returning first few observations

df.head()

#dataset Size df.shape

#List of Column Names df.columns.tolist()

#Data-types of each column df.dtypes

Returning no. of missing values in each column

df.isnull().sum()

df.info()

df.describe()

Replacing 0 by NaN

df[['Glucose','BloodPressure','SkinThickness','Insulin','BMI']] = df[['Glucose','BloodPressure','SkinThickness','Insulin','BMI']].replace(0,np.NaN)

Re-checking no. of missing values in each column

df.isnull().sum()

From the above we found that there are 5 variables having misssing values (Glucose,BloodPressure,SkinThickness,Insulin,BMI)

Filling NaN/Missing values

df['Glucose'].fillna(df['Glucose'].median(), inplace = True) df['BloodPressure'].fillna(df['BloodPressure'].median(), inplace = True) df['SkinThickness'].fillna(df['SkinThickness'].median(), inplace = True) df['Insulin'].fillna(df['Insulin'].median(), inplace = True) df['BMI'].fillna(df['BMI'].mean(), inplace = True)

Checking missing values correctly replaced

df.isnull().sum()

Finding corelation among the dataset

df.corr()

Getting unique values

df['Pregnancies'].unique()

Finding counts of unique values and sorting it in ascending order

df['Pregnancies'].value_counts().sort_values()

#Plotting a graph based on the Outcome Column sns.countplot(x='Outcome', data=df) plt.show()

#Count of Diabetes and Non-Diabetes patients diabetes_count = len(df.loc[df['Outcome'] == 1]) no_diabetes_count=len(df.loc[df['Outcome']==0]) (diabetes_count, no_diabetes_count)

#Plotting graphs for all the parameters except the Outcome ie.,target parameter

cols=['Pregnancies','Glucose','BloodPressure','SkinThickness', 'Insulin','BMI','DiabetesPedigreeFunction','Age'] num=df[cols] for i in num.columns: plt.hist(num[i]) plt.title(i) plt.show()

#here my target column is outcome hence i am splitting the dataset into two where in one the input(X) and the other target(Y) X = df.drop(columns = ['Outcome']) y = df['Outcome']

#Training the dataset X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, stratify=y)

#Performing KNN algorithm for the diabetes dataset knn = KNeighborsClassifier()

#create a dictionary of all values we want to test for n_neighbors params_knn = {'n_neighbors': np.arange(1, 25)}

#useing gridsearch to test all values for n_neighbors knn_gs = GridSearchCV(knn, params_knn, cv=5)

#fit model to training data knn_gs.fit(X_train, y_train)

#save best model knn_best = knn_gs.best_estimator_

#check best n_neigbors value print(knn_gs.best_params_)

#Performing Random Forest Algorithm for the diabetes dataset rf = RandomForestClassifier()

#create a dictionary of all values we want to test for n_estimators params_rf = {'n_estimators': [50, 100, 200]}

#useing gridsearch to test all values for n_estimators rf_gs = GridSearchCV(rf, params_rf, cv=5)

#fit model to training data rf_gs.fit(X_train, y_train)

#save best model rf_best = rf_gs.best_estimator_

#check best n_estimators value print(rf_gs.best_params_)

#Performing logistic regression Algorithm to the diabetes dataset log_reg = LogisticRegression()

#fit the model to the training data log_reg.fit(X_train, y_train)

#Testing the accuracy for all the three models for the diabetes dataset print('knn: {}'.format(knn_best.score(X_test, y_test))) print('rf: {}'.format(rf_best.score(X_test, y_test))) print('log_reg: {}'.format(log_reg.score(X_test, y_test)))

#Performing Ensembly Modelling for the diabetes dataset

#create a dictionary of our models estimators=[('knn', knn_best), ('rf', rf_best), ('log_reg', log_reg)]

#create our voting classifier, inputting our models ensemble = VotingClassifier(estimators, voting='hard')

#fit model to training data ensemble.fit(X_train, y_train)

#test our model on the test data ensemble.score(X_test, y_test)

#Question 1 #Display accuracies for different values of n in KNN Classifier

a_index=list(range(1,11)) a=pd.Series() x=[0,1,2,3,4,5,6,7,8,9,10] for i in list(range(1,11)): model=KNeighborsClassifier(n_neighbors=i) model.fit(X_train,y_train) prediction=model.predict(X_test) a=a.append(pd.Series(metrics.accuracy_score(prediction,y_test))) plt.plot(a_index, a) plt.xticks(x) plt.show() print('Accuracies for different values of n are:',a.values)

Question 2

Display number of rows with respect to the columns

print("Total number of rows : {0}".format(len(df))) print("Number of rows with 0 Pregnancies: {0}".format(len(df.loc[df['Pregnancies'] == 0]))) print("Number of rows with 0 Glucose: {0}".format(len(df.loc[df['Glucose'] == 0]))) print("Number of rows with 0 BloodPressure: {0}".format(len(df.loc[df['BloodPressure'] == 0]))) print("Number of rows with 0 SkinThickness: {0}".format(len(df.loc[df['SkinThickness'] == 0]))) print("Number of rows with 0 Insulin: {0}".format(len(df.loc[df['Insulin'] == 0]))) print("Number of rows with 0 BMI: {0}".format(len(df.loc[df['BMI'] == 0]))) print("Number of rows with 0 DiabetesPedigreeFunction: {0}".format(len(df.loc[df['DiabetesPedigreeFunction'] == 0]))) print("Number of rows with 0 Ages: {0}".format(len(df.loc[df['Age'] == 0])))