Source: View original notebook on GitHub
Category: Machine Learning / Learn ML
Challenge - Diabetes Classification
About Diabetes
Diabetes is a condition that impairs(weakens) the body's ability to process blood glucose, otherwise known as blood sugar. In the United States, the estimated number of people over 18 years of age with diagnosed and undiagnosed diabetes is 30.2 million. The figure represents between 27.9 and 32.7 percent of the population.
Without ongoing, careful management, diabetes can lead to a buildup of sugars in the blood, which can increase the risk of dangerous complications, including stroke and heart disease.
Different kinds of diabetes can occur, and managing the condition depends on the type. Not all forms of diabetes stem from a person being overweight or leading an inactive lifestyle. In fact, some are present from childhood.
In this problem you are given a Diabetes Data set consisting of following features -
['Pregnancies', 'Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI', 'DiabetesPedigreeFunction', 'Age', 'Outcome']
and your task is to predict whether a person is suffering from diabetes or not (Binary Classification)
Tasks
1) Plot a bar graph showing number of classes and no of examples in each class.
2) Classification Task, classify a person as 0 or 1 (Diabetic or Not) using K-Nearest Neighbors classifier.
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
# loading datasets
X = pd.read_csv('Datasets/Assignment4_Diabetes_XTrain.csv')
Y = pd.read_csv('Datasets/Assignment4_Diabetes_YTrain.csv')
test = pd.read_csv('Datasets/Assignment4_Diabetes_Xtest.csv')
# ['Pregnancies', 'Glucose', 'BloodPressure', 'SkinThickness', 'Insulin', 'BMI', 'DiabetesPedigreeFunction', 'Age', 'Outcome']
X.shape
Output:
(576, 8)
Y.shape
Output:
(576, 1)
Y.nunique() # 2 class problem
Output:
Outcome 2
dtype: int64
print(Y.values.flatten())
Output:
[1 0 1 1 0 1 0 0 1 0 0 1 0 1 1 0 1 0 0 0 0 0 0 0 1 0 0 1 0 1 0 0 0 1 0 0 0
0 0 0 0 0 0 0 0 0 0 1 1 1 0 0 1 0 1 0 0 0 0 1 0 1 1 1 0 1 0 1 0 0 0 1 0 0
1 0 0 0 1 0 0 1 0 0 0 1 0 0 1 0 0 1 0 0 0 0 1 1 0 0 0 0 0 0 0 1 0 0 1 0 1
1 1 0 0 0 1 1 1 0 0 1 0 1 1 1 1 0 0 1 0 1 0 0 0 0 0 0 0 0 0 1 1 0 0 1 0 1
1 0 1 1 0 1 1 1 0 0 0 0 0 0 0 0 0 1 0 0 0 1 1 1 1 0 1 0 1 0 0 0 0 0 1 1 1
0 0 0 0 0 0 0 1 1 0 0 0 1 0 0 0 0 0 0 0 1 1 0 1 1 0 0 0 1 0 0 1 1 0 0 1 0
0 0 1 0 0 1 0 1 1 0 1 0 1 1 0 0 1 0 0 0 0 0 1 0 1 0 0 1 0 1 0 1 1 0 0 0 0
0 1 0 0 0 0 1 0 1 0 1 0 1 0 1 0 0 1 1 0 0 0 0 0 0 1 1 0 0 1 0 0 1 0 1 0 0
0 0 0 0 0 1 0 0 1 1 1 1 0 1 0 0 0 1 1 1 0 0 0 0 1 1 1 0 1 0 0 0 0 0 0 0 0
0 0 0 1 0 1 0 1 0 0 0 1 0 0 0 1 1 0 0 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0
1 0 1 1 1 0 1 1 1 0 0 0 0 1 0 0 1 1 0 1 0 0 0 0 1 0 0 0 0 0 0 1 1 0 0 1 0
1 0 1 0 0 0 1 0 0 1 1 0 1 1 1 1 1 0 0 0 0 1 0 0 0 0 0 1 0 1 0 0 0 0 0 1 1
0 0 0 1 1 0 0 0 0 1 0 0 0 0 0 0 1 0 0 1 0 0 0 0 0 1 0 0 0 0 0 0 0 0 1 1 1
0 1 1 0 0 0 0 0 1 1 0 0 1 0 0 1 0 0 1 1 1 0 1 1 0 0 0 0 0 0 1 1 0 0 1 0 1
0 1 0 0 0 1 0 1 0 1 0 0 1 1 0 0 0 0 0 0 1 0 1 0 0 0 1 1 0 0 1 0 0 0 0 1 1
1 1 0 0 0 1 0 0 1 0 0 0 1 0 0 0 0 1 0 0 0]
# part 1) Plot a bar graph showing number of classes and no of examples in each class.
values, count = np.unique(Y,return_counts=True)
values
Output:
array([0, 1], dtype=int64)
count
Output:
array([375, 201])
plt.bar(values,count)
plt.xticks([0,1])
plt.yticks(count)
plt.title('Part 1 bargraph ')
plt.xlabel('Classes')
plt.ylabel('Freuency')
Output:
Text(0, 0.5, 'Freuency')
Y.head()
Output:
Outcome
0 1
1 0
2 1
3 1
4 0
X.columns
Output:
Index(['Pregnancies', 'Glucose', 'BloodPressure', 'SkinThickness', 'Insulin',
'BMI', 'DiabetesPedigreeFunction', 'Age'],
dtype='object')
part 2) Classification Task, classify a person as 0 or 1 (Diabetic or Not).
# using Knn code written from scratch
def distance(x1,x2):
# x1 and x2 are both numpy arrays
inside_part = np.sum((x1-x2)**2)
return np.sqrt(inside_part)
def Knn(X, Y, querypoint, K=5) :
# X and Y are numpy arrays
Y = Y.flatten()
distance_label_pair = []
for i in range(X.shape[0]):
dist = distance(querypoint, X[i])
distance_label_pair.append((dist,Y[i]))
# sorting dictionary based upon distance
distance_label_pair.sort() # sorting will be done based on 1st values of pair of list
# k-nearest
k_nearest = distance_label_pair[:K]
ans = []
for dist,label in k_nearest :
ans.append(label)
values,counts = np.unique(ans, return_counts=True)
max_index = np.argmax(counts)
return values[max_index]
from sklearn.model_selection import train_test_split
X_train, X_test, Y_train, Y_test = train_test_split(X,Y, test_size=0.25, random_state = 101)
# prediction over X_test
Y_pred = []
for query in X_test.values :
Y_pred.append(Knn(X_train.values,Y_train.values, query , K=100))
Y_pred = np.array(Y_pred)
Y_pred
Output:
array([0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 0,
0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 0, 0, 0,
1, 0, 1, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0,
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1], dtype=int64)
Y_test.values.flatten()
Output:
array([1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 1, 0,
1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 1, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0,
0, 0, 1, 0, 1, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1, 0, 1, 0, 0, 0,
1, 1, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0,
0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,
0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1], dtype=int64)
accuracy = np.sum(Y_pred == Y_test.values.flatten()) / Y_pred.shape[0]
accuracy
Output:
0.7222222222222222
# output
# prediction over test
output = []
for query in test.values :
output.append(Knn(X_train.values,Y_train.values, query , K=100))
output = np.array(output)
output
Output:
array([0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0,
1, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1,
0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0,
1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 1, 1, 1, 0, 1, 0, 0, 0,
0, 0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0,
0, 0, 0, 1, 1, 0, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0], dtype=int64)
df = pd.DataFrame(output,columns=['Outcome'])
df.head()
Output:
Outcome
0 0
1 0
2 0
3 0
4 0
df.to_csv('Datasets/Assignment4_Diabetes_output_scratch.csv',index = False)
pd.read_csv('Datasets/Assignment4_Diabetes_output_scratch.csv').head() # got 70% accuracy when submitted
Output:
Outcome
0 0
1 0
2 0
3 0
4 0
part 2) Classification Task, classify a person as 0 or 1 (Diabetic or Not) using K-Nearest Neighbors classifier.
Using Sklearn(sklearn.neighbors.KNeighborsClassifier)
from sklearn.preprocessing import StandardScaler
s = StandardScaler()
X = pd.DataFrame(s.fit_transform(X))
test = pd.DataFrame(s.transform(test))
Output:
c:\users\shaurya singhal\appdata\local\programs\python\python37-32\lib\site-packages\sklearn\preprocessing\data.py:625: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by StandardScaler.
return self.partial_fit(X, y)
c:\users\shaurya singhal\appdata\local\programs\python\python37-32\lib\site-packages\sklearn\base.py:462: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by StandardScaler.
return self.fit(X, **fit_params).transform(X)
c:\users\shaurya singhal\appdata\local\programs\python\python37-32\lib\site-packages\ipykernel_launcher.py:4: DataConversionWarning: Data with input dtype int64, float64 were all converted to float64 by StandardScaler.
after removing the cwd from sys.path.
from sklearn.model_selection import train_test_split
X_train, X_test, Y_train, Y_test = train_test_split(X,Y, test_size=0.25, random_state = 101)
from sklearn.neighbors import KNeighborsClassifier
model = KNeighborsClassifier(n_neighbors=100)
X_train.shape
Output:
(432, 8)
Y_train.shape
Output:
(432, 1)
model.fit(X_train.values,Y_train.values.reshape(-1))
Output:
KNeighborsClassifier(algorithm='auto', leaf_size=30, metric='minkowski',
metric_params=None, n_jobs=None, n_neighbors=100, p=2,
weights='uniform')
model.get_params()
Output:
{'algorithm': 'auto',
'leaf_size': 30,
'metric': 'minkowski',
'metric_params': None,
'n_jobs': None,
'n_neighbors': 100,
'p': 2,
'weights': 'uniform'}
model.predict(test)
Output:
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0,
0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0,
1, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0,
0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0,
0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int64)
model.score(X_test,Y_test)
Output:
0.7361111111111112
# output
output = model.predict(test)
output
Output:
array([0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 1, 1, 1, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0,
0, 0, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0,
1, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0,
0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0,
0, 1, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0], dtype=int64)
type(output)
Output:
numpy.ndarray
df = pd.DataFrame(output,columns=['Outcome'])
df.head()
Output:
Outcome
0 0
1 0
2 0
3 0
4 0
df.to_csv('Datasets/Assignment4_Diabetes_output.csv',index = False)
pd.read_csv('Datasets/Assignment4_Diabetes_output.csv').head()
Output:
Outcome
0 0
1 0
2 0
3 0
4 0