Skip to contents

Downloads a tsf file from Zenodo using the provided record ID and dataset name.

Usage

download_zenodo_record(record_id = 4656222, dataset_name = "m3_yearly_dataset")

Arguments

record_id

(character(1)) the Zenodo record ID.

dataset_name

(character(1)) the name of the dataset to download.

References

Godahewa, Rakshitha, Bergmeir, Christoph, Webb, I G, Hyndman, J R, Montero-Manso, Pablo (2021). “Monash time series forecasting archive.” arXiv preprint arXiv:2105.06643.

Examples

if (FALSE) { # \dontrun{
dt = download_zenodo_record(record_id = 4656222, dataset_name = "m3_yearly_dataset")

# optional renaming
setnames(dt, c("id", "date", "value"))

# transform into single task
task = as_task_fcst(dt)

# or split up for forecast learners that don't allow key columns
tasks = split(dt, ~id) |> map(remove_named, "id") |> map(as_task_fcst)
} # }