Tblock_SPcontig: Temporal blocked and contiguously-blocked spatial CV

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

Description

Fold allocation of k-fold CV using:

• blocked time

• contiguously blocked locations

Usage

Tblock_SPcontig(data, nfolds, t.nfolds = round(sqrt(nfolds)),
  sp.nfolds = round(sqrt(nfolds)), time = "time", site_id = "site")

Arguments

data	full dataset
nfolds	number of folds
t.nfolds	number of folds across time. Default is sqrt(nfolds)
sp.nfolds	number of folds across space. Default is sqrt(nfolds)
time	column name of time-stamp in data. Default is "time"
site_id	column name of location identifier in data. Default is "site"

Value

A list with slots:

• data, possibly re-ordered

• f, a vector with the fold identifiers of each row in data. The fold identifier is composed of the concatenation of time-fold number (from 1 to t.nfolds) and space-fold number (from 1 to sp.nfolds), separated by "_".

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