import pandas as pd
import json
import numpy as np
import matplotlib.pyplot as plt
from PIL import Image
from sklearn.model_selection import train_test_split

from keras.models import Sequential
from keras.layers import Dense
import numpy
# fix random seed for reproducibility
numpy.random.seed(7)



with open("train.json") as f:
    d=json.loads(f.read())

df=pd.DataFrame(d)
df=df.drop(["id","band_2"],axis=1)
df.info()
data=[]
for i in range(1604):
    data1=df.ix[i,0]
    dataf=np.reshape(data1,(75,75))
    dataf2=dataf[20:50,20:50]
    dataf3=np.reshape(dataf2,(900,))
    data.append(dataf3)

data=np.array(data)
"""
names=[i for i in range(1604)]

X=pd.DataFrame()
for i in range(1604):
    data1=df.ix[i,0]
    dataf=np.reshape(data1,(75,75))
    FD=dataf[20:50,20:50].reshape(900,1)
    for j in range(900):
        X.ix[i,j]=FD[j]


data=[]
for i in range(1604):

    data.append(df.ix[i,0])

"""



y=df.ix[:,"is_iceberg"]


X_train, X_test, y_train, y_test = train_test_split(data, y, test_size=0.33, random_state=42)
import seaborn as sns
from sklearn.neighbors import KNeighborsClassifier
from sklearn.metrics import confusion_matrix
neigh = KNeighborsClassifier(n_neighbors=3)
from sklearn.model_selection import cross_val_score
neigh.fit(X_train, y_train)
y_predict=neigh.predict(X_test)

tn, fp, fn, tp = confusion_matrix(y_test, y_predict, labels=[0,1]).ravel()
print((tn, fp, fn, tp))
accuracy=100*(tp+fn)/len(y_predict)
print(accuracy)
sns.set()


scores = cross_val_score(neigh, X, y, cv=10)
print(100*np.mean(scores))