run_all_experiments: Run multiple experiments for different grid and time series...

In mrfoliveira/Evaluation-procedures-for-forecasting-with-spatio-temporal-data: Evaluation procedures for forecasting with spatio-temporal data

Description

Run multiple experiments for different grid and time series sizes

Usage

run_all_experiments(models, nested_data_list, in_set_perc, form,
  in_estimators, in_est.pars, out_estimator = "t_oos",
  out_est.pars = list(tr.perc = in_set_perc),
  workflow = "simple_workflow", wf.pars = NULL,
  evaluator = "evaluate", eval.pars = NULL, seed = 1234,
  site_id = "site", time = "time", .compress = FALSE,
  .progress = NULL, .verbose = TRUE, .saveMem = FALSE)

Arguments

models	vector of names of models to use for learning
nested_data_list	a nested list of data sets: top level identifies grid size, second level identifies time series size, and third level contains lists of multiple data sets
in_set_perc	a fraction of the data to be used as in-set
form	a learning formula
in_estimators	a vector of names of estimator functions to use on the in-set data
in_est.pars	a named list of the same length as in_estimators containing lists of arguments to feed to each estimator apllied to the in-set data
out_estimator	the name of the estimator function to use on the out-set data
out_est.pars	a list containing arguments to feed to the estimator apllied to the out-set data
workflow	the name of the function implementing a workflow
wf.pars	a list of arguments to feed to workflow
evaluator	the name of the function to calculate evaluation metrics
eval.pars	a list of arguments to feed to evaluator
seed	a seed to set at the start of the experiment
site_id	the name of the data column containing location identifiers
time	the name of the data column containing time-stamps
.compress	a Boolean indicating whether to compress results
.progress	a file name to save temporary results
.verbose	a Boolean indicating whether progress should be reported on during experiments
.saveMem	a Boolean indicating whether partial results should be discarded to save memory. Only possible if .progress is TRUE. Default is FALSE.

Value

A nested list: top top level identifies grid size, second level identifies time series size, and third level contains lists with two slots: out_estRes containing the results of the out-set estimators, and in_estRes containing a list of results of each estimator used on in-set data

See Also

run_multiple_experiments

mrfoliveira/Evaluation-procedures-for-forecasting-with-spatio-temporal-data documentation built on April 11, 2021, 10:50 a.m.