model.py

'''







'''
import pandas as pd
import numpy as np
from sklearn.model_selection import train_test_split, learning_curve
from sklearn.metrics import average_precision_score
from sklearn.linear_model import LogisticRegression
import pickle

df = pd.read_csv('Fraud_data_kaggle.csv')
X = df.loc[(df.type == 'TRANSFER') | (df.type == 'CASH_OUT')]

randomState = 5
np.random.seed(randomState)

#X = X.loc[np.random.choice(X.index, 100000, replace = False)]

Y = X['isFraud']
del X['isFraud']

# Eliminate columns shown to be irrelevant for analysis in the EDA
X = X.drop(['nameOrig', 'nameDest', 'isFlaggedFraud'], axis = 1)

# Binary-encoding of labelled data in 'type'
X.loc[X.type == 'TRANSFER', 'type'] = 0
X.loc[X.type == 'CASH_OUT', 'type'] = 1
X.type = X.type.astype(int) # convert dtype('O') to dtype(int)

Xfraud = X.loc[Y == 1]
XnonFraud = X.loc[Y == 0]
print('\nThe fraction of fraudulent transactions with \'oldBalanceDest\' = \
\'newBalanceDest\' = 0 although the transacted \'amount\' is non-zero is: {}'.\
format(len(Xfraud.loc[(Xfraud.oldbalanceDest == 0) & \
(Xfraud.newbalanceDest == 0) & (Xfraud.amount)]) / (1.0 * len(Xfraud))))

print('\nThe fraction of genuine transactions with \'oldBalanceDest\' = \
newBalanceDest\' = 0 although the transacted \'amount\' is non-zero is: {}'.\
format(len(XnonFraud.loc[(XnonFraud.oldbalanceDest == 0) & \
(XnonFraud.newbalanceDest == 0) & (XnonFraud.amount)]) / (1.0 * len(XnonFraud))))

X.loc[(X.oldbalanceDest == 0) & (X.newbalanceDest == 0) & (X.amount != 0), \
      ['oldbalanceDest', 'newbalanceDest']] = - 1
      
      
X.loc[(X.oldbalanceDest == 0) & (X.newbalanceDest == 0) & (X.amount != 0), \
      ['oldbalanceDest', 'newbalanceDest']] = np.nan      

X['errorBalanceOrig'] = X.oldbalanceOrg + X.amount - X.oldbalanceOrg
X['errorBalanceDest'] = X.oldbalanceDest + X.amount - X.newbalanceDest

trainX, testX, trainY, testY = train_test_split(X, Y, test_size = 0.2, \
                                                random_state = randomState)

# Long computation in this cell (~1.8 minutes)
# weights = (Y == 0).sum() / (1.0 * (Y == 1).sum())
clf = LogisticRegression()
log = clf.fit(trainX, trainY)
predict = log.predict(testX)  

pickle.dump(log, open('logistic.pkl', 'wb'))