forecast_all: Forecast all requested models

In bdemeshev/torro:

Description

forecast_all forecasts all requested models.

Usage

forecast_all(fits_long, var_sets, n_cores = 1, basefolder,
  log_file = "estim_log.txt", ...)

Arguments

fits_long	data frame with requested models
var_sets	correspondance between variable sets and variable names
n_cores	number of cores used. Just ignored.
basefolder	path to folder where results are stored
log_file	the name of log file
...	further arguments passet to forecast_one_fit

Details

Just do it. This function may take a long time. If you stop it, you may recover estimation status (column 'result' in fits_long) from written fit files using get_status_from_fit_files(). It is better to clear basefolder before estimation process. The function will just overwrite old '.Rds' files in case of name clash.

Value

fits_long with estimation messages for each requested model.

Examples

fits_long_toy <- create_fits_long(shifts_toy, models_toy, var_sets_toy, horizons_toy)
fits_long_very_toy <- fits_long_toy[1:2, ]
res <- forecast_all(fits_long_very_toy, var_sets_toy, n_cores = 1,
    basefolder = tempdir())

bdemeshev/torro documentation built on May 21, 2019, 8:36 a.m.