knn.py

#!/usr/bin/env python3

#####################################################
##    WISCONSIN BREAST CANCER MACHINE LEARNING     ##
#####################################################

"""
Kth Nearest Neighbor Classification
"""

import time
import sys, os
import pandas as pd
import helper_functions as hf
from helper_functions import training_set, class_set, test_set, test_class_set
from sklearn.neighbors import KNeighborsClassifier # Kth Nearest Neighbor
from sklearn.model_selection import KFold, cross_val_score # Cross validation
from sklearn.model_selection import KFold, GridSearchCV
from sklearn.metrics import roc_curve # ROC Curves
from sklearn.metrics import auc # Calculating Area Under Curve for ROC's!
from sklearn.externals import joblib

# Fitting model
fit_knn = KNeighborsClassifier(n_neighbors=3)

# Training model
fit_knn.fit(training_set, 
	class_set['diagnosis'])

# Since KNN was first algorithm I included training set metrics
# to give context  
# We predict the class for our training set
predictionsTrain = fit_knn.predict(training_set) 
	
# Measure the accuracy based on the training set
accuracy_train = fit_knn.score(training_set, 
	class_set['diagnosis'])

train_error_rate = 1 - accuracy_train  

# First we predict the Dx for the test set and call it predictions
predictions = fit_knn.predict(test_set)	

# Let's get the accuracy of our test set
accuracy = fit_knn.score(test_set, 
	test_class_set['diagnosis'])

test_error_rate = 1 - accuracy

# ROC Curve and AUC Calculations
fpr, tpr, _ = roc_curve(predictions, 
	test_class_set)

auc_knn = auc(fpr, tpr)

# Uncomment to save your model as a pickle object!
# joblib.dump(fit_knn, 'pickle_models/model_knn.pkl')

if __name__ == '__main__':
	print('''
		#################################
		## FITTING MODEL KNN USING k=3 ##
		#################################
		'''
		)
		
	print(fit_knn)
	
	print('''
		###############
		## Optimal K ##
		###############
			''')
	# KNN Optimal K
	# Inspired by: 
	# https://kevinzakka.github.io/2016/07/13/k-nearest-neighbor/
		
	myKs = []
	for i in range(0, 50):
		if (i % 2 != 0):
			myKs.append(i)	

	cross_vals = []
	for k in myKs:
		knn = KNeighborsClassifier(n_neighbors=k)
		scores = cross_val_score(knn,
			training_set, 
			class_set['diagnosis'], 
			cv = 10, 
			scoring='accuracy')
		cross_vals.append(scores.mean())

	MSE = [1 - x for x in cross_vals]
	optimal_k = myKs[MSE.index(min(MSE))]
	print("Optimal K is {0}".format(optimal_k))
		
	print('''
		###############################
		## TRAINING SET CALCULATIONS ##
		###############################
		'''
		)
	
	# Here we create a matrix comparing the actual values 
	# vs. the predicted values
	print(pd.crosstab(predictionsTrain, 
		class_set['diagnosis'], 
		rownames=['Predicted Values'], 
		colnames=['Actual Values']))
		
	print("Here is our accuracy for our training set:\n {0: .3f}"\
		.format(accuracy_train))
	
	print("The train error rate for our model is:\n {0: .3f}"\
		.format(train_error_rate))
	
	print('''
		###############################
		##      CROSS VALIDATION     ##
		###############################
		'''
		)
	
	hf.cross_val_metrics(fit_knn, training_set, class_set['diagnosis'], 
		print_results = True)
			
	print('''
		###############################
		##   TEST SET CALCULATIONS   ##
		###############################
		'''
		)
		
	# Let's compare the predictions vs. the actual values
	print(pd.crosstab(predictions, 
		test_class_set['diagnosis'], 
		rownames=['Predicted Values'], 
		colnames=['Actual Values']))
		
	# TEST ERROR RATE!!
	print("Here is our accuracy for our test set:\n {0: .3f}"\
		.format(accuracy))
	
	# Here we calculate the test error rate!
	print("The test error rate for our model is:\n {0: .3f}"\
		.format(test_error_rate))

	# ROC Curve
	# NOTE: These functions were created in the helperFunctions.py 
	# script to reduce lines of code
	# refer to helper.py for additional information
	hf.plot_roc_curve(fpr, tpr, auc_knn, 'knn')
	
	# Zoomed in ROC Curve
	hf.plot_roc_curve(fpr, tpr, auc_knn, 'knn',
		(-0.001, 0.2), (0.7, 1.05))
else:
	def return_knn():
		'''
		Function to output values created in script 
		'''
		return fpr, tpr, auc_knn, predictions, test_error_rate

	mean_cv_knn, std_error_knn = hf.cross_val_metrics(fit_knn, 
		training_set, 
		class_set['diagnosis'],
		print_results = False)