def train(self):
print(path)
if self.path == 'data/bank-additional-full.csv':
data = pd.read_csv('data/bank-additional-full.csv', sep=';')
data['y'] = data['y'].map({'yes': 1, 'no': 0})
else:
data = pd.read_csv('path')
process = Preprocessor(data)
data, columns = process._divide_data(data, self.label)
categorical, numerical = process._classify_data(columns, data)
transformed, y = process._preprocess_data(categorical,
numerical,
preprocessor=scaler)
#dealing with our imbalanced data by oversampling the data set
model = Modelling()
x_train, x_test, y_train, y_test = model._splitdata(transformed,
y,
size=self.validate)
X, y = DealwithSample(x_train, y_train, method=sampler)
model = model.Prediction(X, x_test, y, y_test, method=self.model)
