#!/usr/bin/env python3
import pandas as pd
import numpy as np
import datetime
import sys

from sklearn.model_selection import train_test_split
from sklearn.metrics import roc_auc_score
from sklearn.metrics import accuracy_score
from sklearn.metrics import precision_score
from sklearn.metrics import recall_score
from sklearn.metrics import f1_score
from sklearn.metrics import confusion_matrix

from sklearn.neighbors import KNeighborsClassifier
from sklearn.linear_model import * 
from sklearn.ensemble import RandomForestClassifier
from sklearn.ensemble import GradientBoostingClassifier
from sklearn.preprocessing import StandardScaler
from sklearn.svm import SVC
from sklearn.neural_network import MLPClassifier


#       load train data 
data=pd.read_csv('traindata_sea.csv')

#       add feature colomn
data['flash'] = np.heaviside(data['lightning_count']-1,0)

#       select target colomn
wwlln=data['flash']      

#       del all but sattelite data
del data['flash']
del data['lat']
del data['lon']
del data['ptime']
del data['lightning_count']
del data['avg_energy']
print(data.head())

folds=5         # how many folds
scaler = StandardScaler()

#       select classificator
clf = LogisticRegression(penalty='l2', class_weight={1: 0.774})
#clf = RandomForestClassifier(n_estimators=50)
#clf = MLPClassifier(hidden_layer_sizes=(5,), shuffle=True,verbose=False)

accur, recall, f1 = 0, 0 ,0
matrix = np.zeros((4))
start_time = datetime.datetime.now()
coefs = np.empty([1,data.shape[1]])

for fold in range(folds):
    print('N_fold is:', fold)
    X_train, X_test, y_train, y_test = train_test_split(data, wwlln, test_size=1/folds, shuffle=True)
    X_train_scaled = scaler.fit_transform(X_train)
    X_test_scaled = scaler.transform(X_test)

    clf.fit(X_train_scaled, y_train)

    try:
        coefs += clf.coef_
    except: 
        print('There are NO coeffs for this clf!')

    predict = clf.predict(X_test_scaled)
    accur += precision_score(y_test, predict)
    recall += recall_score(y_test, predict)
    f1 += f1_score(y_test, predict)    
    matrix += confusion_matrix(y_test, predict).ravel()
    print (f1)
    time = datetime.datetime.now() - start_time

print (pd.DataFrame((coefs/folds)/np.max(np.abs(coefs/folds))))
print ('accur', accur/folds)
print ('recall', recall/folds)
print ('f1', f1/folds)

print ('TN', 'FP', 'FN', 'TP')
print (matrix)
print('Total events:', matrix.sum())
print('By set:', 105295)
if matrix.sum() == 105295:
    print('Match! OK!')

print ('time', time)