classifier.py

# importing required libraries
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
# import GaussianNB
from sklearn.naive_bayes import GaussianNB
import cv2
# import warnings to remove any type of future warnings
import warnings
warnings.simplefilter(action='ignore', category=FutureWarning)

# reading csv file and extracting class column to y.
dataf = pd.read_csv("Datasetinfectedhealthy.csv")

# extracting two features
X = dataf.drop(['imgid','fold num'], axis=1)
y = X['label']
X = X.drop('label', axis=1)

print("\nTraining dataset:-\n")
print(X)


log = pd.read_csv("datasetlog/Datasetunlabelledlog.csv")

log = log.tail(1)
X_ul = log.drop(['imgid','fold num'], axis=1)

print("\nTest dataset:-\n")
print(X_ul)

print("\n*To terminate press (q)*")

#X.plot(kind='scatter',x='feature1',y='feature2')
#plt.show()


Sum = 0

from sklearn.model_selection import train_test_split

'''
from sklearn.model_selection import train_test_split  
for n in range(4):
	x_train, Xi_test, y_train, yi_test = train_test_split(X, y, test_size=0.52, random_state=60)  
	if cv2.waitKey(1) == ord('q' or 'Q'): break     
	svclassifier = SVC(kernel='linear')  
	svclassifier.fit(x_train, y_train)  
	pred = svclassifier.predict(X_ul)
'''
for n in range(4):
	x_train, Xi_test, y_train, yi_test = train_test_split(X, y, test_size=0.52, random_state=60)
	if cv2.waitKey(1) == ord('q' or 'Q'): break
	classifier = GaussianNB()
	classifier.fit(x_train, y_train)
	pred =classifier.predict(X_ul)
	Sum = Sum + pred
	print(pred)

print("\nprediction: %d" %int(Sum/4))

if(Sum < 2):
	print("The leaf is sufficiently healthy!")
else:
	print("The leaf is infected!")

print("\nKeypress on any image window to terminate")

#from sklearn.metrics import classification_report, confusion_matrix

#print(classification_report(yi_test,y_pred))
#print "\n Average precision percentage: %.2f"  %avg_pred + "%"
cv2.waitKey(0)