RemixAutoML (version 0.11.0)

ID_Forecast: ID_Forecast for forecasting intermittent demand data

Description

ID_Forecast for forecasting intermittent demand data. It runs collapsed gibbs sampler simulations, using the count and size quantile regression models, for every period for all group levels in your data.

Usage

ID_Forecast(CountData = FinalData$CountData,
  SizeData = FinalData$SizeData,
  CountDataNames = FinalData$CountPredNames,
  SizeDataNames = FinalData$SizePredNames, GroupVar = NULL,
  FC_Periods = 26, NumSims = 1000, PredictionIntervals = c(0.05, 0.2,
  0.8, 0.95))

Arguments

CountData

This is the count data returned from AutoCatBoostFreqSizeScoring() or AutoH2oGBMFreqSizeScoring()

SizeData

This is the size data returned from AutoCatBoostFreqSizeScoring() or AutoH2oGBMFreqSizeScoring()

CountDataNames

This is the count data names returned from AutoCatBoostFreqSizeScoring() or AutoH2oGBMFreqSizeScoring()

SizeDataNames

This is the size data returned from AutoCatBoostFreqSizeScoring() or AutoH2oGBMFreqSizeScoring()

GroupVar

This is your grouping variable. E.g. sku as being the name of the column that contains all skus

FC_Periods

The max period of your forecast. E.g. if you want 52 weeks of forecasts, set to 52

NumSims

Set the number of collapsed gibbs simulations to run for each time unit forecast. E.g. if you want 52 weeks of forecasts, the simulations will run NumSims for each period up to 52.

PredictionIntervals

Set the prediction intervals you want returned. E.g. c(seq(0.05,0.95,0.05))

Value

Returns your entire set of groupvar forecasts in a single data.table

See Also

Other Automated Time Series: AutoCatBoostCARMA, AutoCatBoostFreqSizeScoring, AutoCatBoostSizeFreqDist, AutoH2oDRFCARMA, AutoH2oGBMCARMA, AutoH2oGBMFreqSizeScoring, AutoH2oGBMSizeFreqDist, AutoTS, AutoXGBoostCARMA, ID_SingleLevelGibbsSampler, IntermittentDemandScoringDataGenerator

Examples

# NOT RUN {
Results <- ID_Forecast(
  CountData, 
  SizeData, 
  GroupVar = "sku", 
  FC_Periods = 26, 
  NumSims = 1000, 
  PredictionIntervals = c(0.05,0.20,0.50,0.80,0.95))
# }