def get_accuracy(model_type, stock_name, start, stop, p_days, period,
dist_period):
"""
:param model_type:
:param stock_name:
:param start:
:param stop:
:param p_days:
:param period:
:param dist_period:
:return:
"""
rw_data = get_raw_data(stock_name, start, stop)
ft_data = add_feat(rw_data, p_days)
new_ft_data = add_bucket(ft_data, dist_period)
X_features = new_ft_data.columns.tolist()
y_index = X_features.index('Tmrw_return')
del X_features[y_index]
z_index = X_features.index('expect_ret')
del X_features[z_index]
new_ft_data['Ticker'] = stock_name
if model_type == "rf":
t_index = X_features.index(yclass_label)
del X_features[t_index]
input_X = new_ft_data[X_features]
input_Y = new_ft_data[yclass_label]
accuracy, pred_table = fitting_forest(input_X,
input_Y,
period,
n_estimators=100)
return 100 * accuracy
pred_table = calc_pred_data(new_ft_data, period, X_features)
acc = acc_pred(pred_table)
return acc
return pred_table
if __name__ == "__main__":
stock_name = sys.argv[1]
start = sys.argv[2]
stop = sys.argv[3]
p_days = int(sys.argv[4])
period = int(sys.argv[5])
dist_period = int(sys.argv[6])
check_data_input(p_days, period, dist_period, start, stop)
# Getting Raw data
rw_data = get_raw_data(stock_name, start, stop)
# Formating data to add 4*p_days features
ft_data = add_feat(rw_data, p_days)
# New format to add bucket
new_ft_data = add_bucket(ft_data, dist_period)
# Getting the feat array
X_features = new_ft_data.columns.tolist()
y_index = X_features.index('Tmrw_return')
del X_features[y_index]
z_index = X_features.index('expect_ret')
del X_features[z_index]
new_ft_data['Ticker'] = stock_name