bsts_create: Create a bsts_model

In lissahyacinth/bstsTest: Rapidly Iterate & Test BSTS Models

Description

Create a bsts_model

Usage

bsts_create(df, date_variable, response, nseasons, niter = 100,
  target_variable, group_variable, model.options = BstsOptions(),
  rebag_vars = FALSE, rebag_mean_vars = FALSE,
  inclusion_probability = 0.01)

Arguments

df	Dataframe
date_variable	Column name containing date
response	Group that is being forecast
nseasons	Seasonality of data - will be assumed if not provided
niter	Number of MCMC iterations to attempt
target_variable	Column name containing metric to be forecast
group_variable	Column name indicating group names - will build into regressors
model.options	Additional options for BSTS
rebag_vars	Create pseudo aggregate variable of other regressors
rebag_mean_vars	Create psuedo mean variable of other regressors
inclusion_probability	Minimum probability of inclusion in final model to show in returned predictors

Value

list object - bsts.model, data.frame of predictors

lissahyacinth/bstsTest documentation built on May 8, 2019, 11:19 p.m.