croston

croston model for intermittent time series

croston.croston.fit_croston(input_endog, forecast_length, croston_variant='original')

Parameters:

• input_endog – numpy array of intermittent demand time series
• forecast_length – forecast horizon
• croston_variant – croston model type

Returns:

dictionary of model parameters, in-sample forecast, and out-of-sample forecast

croston.croston._croston(input_series, input_series_length, croston_variant, w, h, epsilon)

croston.croston._croston_opt(input_series, input_series_length, croston_variant, epsilon, w=None, nop=1)

croston.croston._croston_cost(p0, input_series, input_series_length, croston_variant, epsilon)