import pandas as pd
from sklearn.preprocessing import LabelEncoder
from sklearn.model_selection import train_test_split
from keras.models import Sequential
from keras.layers import Dense

from input_categorical import InputCategorical


df = pd.read_csv('train.csv')
X = df.drop('ACTION', axis=1)
y = df['ACTION']

# Encode categorical features (all 9 features) as ordinal
le = LabelEncoder()
for col in X:
    X[col] = le.fit_transform(X[col])

X = X.values
y = y.values

model = Sequential()
model.add(InputCategorical(input_dim=9,
                           categorical=[0, 1, 2, 3, 4, 5, 6, 7, 8],
                           max_level=[X[:, 0].max(),
                                      X[:, 1].max(),
                                      X[:, 2].max(),
                                      X[:, 3].max(),
                                      X[:, 4].max(),
                                      X[:, 5].max(),
                                      X[:, 6].max(),
                                      X[:, 7].max(),
                                      X[:, 8].max()],
                           embedding_dim='4throot'
                           ))
model.add(Dense(32, activation='relu'))
model.add(Dense(1, activation='sigmoid'))

model.compile(loss='binary_crossentropy', optimizer='sgd', metrics=['accuracy'])

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)

model.fit(X_train, y_train)
score = model.evaluate(X_test, y_test)