classification.py

import os
import os.path
import glob
from keras.models import model_from_json
from keras.layers import Dense
from keras.optimizers import SGD
import cv2
import numpy as np
import pickle as pkl
import sys

from sklearn.preprocessing import MinMaxScaler

from sklearn.naive_bayes import GaussianNB

def get_data_parts(dataline):
        parts = dataline.split(" ")
        Y = np.array([int(parts[0]),int(parts[1]),int(parts[2]),int(parts[3]),int(parts[4]),int(parts[5]),int(parts[6]),int(parts[7]),int(parts[8])])
        Y = Y.reshape(9,1)
        y_train = np.zeros([1])
        #print(y_train)
        if int(parts[9]) == 1:
                y_train[0] = float(0.1 * 10)
        else :
                y_train[0] = float(0)
        #y_train[0][int(parts[10])] = 0.95/4.5
        #y_train[0][int(parts[11])] = 0.9/4.5
        #y_train[0][int(parts[12])] = 0.8/4.5
        #y_train[0][int(parts[13])] = 0.85/4.5
        #print(y_train[0][int(parts[9])])
        return Y , y_train
        
if __name__ == '__main__':
        num_classes = 21
        f1  = open('datasettest.txt','r')
        n = f1.readline().strip()
        k = 0
        while k != 20000:
                Y , y_train = get_data_parts(n)
                for i in range(19999) :
                        n = f1.readline().strip()
                        z , z_train = get_data_parts(n)
                        #print(np.shape(Y))
                        Y = np.append(Y,z,axis = 1)
                        y_train = np.append(y_train,z_train,axis = 0)
                        k+=1
                n = f1.readline().strip()
                k+=1
        Y = Y.T
        scalar = MinMaxScaler()
        Y = scalar.fit_transform(Y)
        
        gnb = GaussianNB()
        gnb.fit(Y,y_train)
        print(gnb.score(Y,y_train))
        #print(Y.shape)
        #print(y_train.shape)
        f2  = open('datasetclass.txt','r')
        n = f2.readline().strip()
        k = 0
        while k != 100000:
                Y_test , y_test = get_data_parts(n)
                for i in range(99999) :
                        n = f2.readline().strip()
                        z , z_train = get_data_parts(n)
                        #print(np.shape(Y))
                        Y_test = np.append(Y_test,z,axis = 1)
                        y_test = np.append(y_test,z_train,axis = 0)
                        k+=1    
                n = f2.readline().strip()
                k+=1
        Y_test = Y_test.T
        Y_test = scalar.fit_transform(Y_test)
        print(gnb.score(Y_test,y_test))