surveyCV (version 0.2.0)

folds.svy: Creating CV folds based on the survey design

Description

This function creates a fold ID for each row in the dataset, to be used for carrying out cross validation on survey samples taken using a SRS, stratified, clustered, or clustered-and-stratified sampling design. Returns a vector of fold IDs, which in most cases you will want to append to your dataset using cbind or similar (see Examples below). These fold IDs respect any stratification or clustering in the survey design. You can then carry out K-fold CV as usual, taking care to also use the survey design features and survey weights when fitting models in each training set and also when evaluating models against each test set.

Usage

folds.svy(Data, nfolds, strataID = NULL, clusterID = NULL)

Arguments

Data

Dataframe of dataset

nfolds

Number of folds to be used during cross validation

strataID

String of the variable name used to stratify during sampling, must be the same as in the dataset used

clusterID

String of the variable name used to cluster during sampling, must be the same as in the dataset used

Value

Integer vector of fold IDs with length nrow(Data). Most likely you will want to append the returned vector to your dataset, for instance with cbind (see Examples below).

Details

If you have already created a svydesign object, you will probably prefer the convenience wrapper function folds.svydesign.

For the special cases of linear or logistic GLMs, use instead cv.svy, cv.svydesign, or cv.svyglm which will automate the whole CV process for you.

See Also

folds.svydesign for a wrapper to use with a svydesign object

cv.svy, cv.svydesign, or cv.svyglm to carry out the whole CV process (not just forming folds but also training and testing your models) for linear or logistic regression models

Examples

# NOT RUN {
# Set up CV folds for a stratified sample and a one-stage cluster sample,
# using data from the `survey` package
library(survey)
data("api", package = "survey")
# stratified sample
apistrat <- cbind(apistrat,
                  .foldID = folds.svy(apistrat, nfolds = 5, strataID = "stype"))
# Each fold will have observations from every stratum
with(apistrat, table(stype, .foldID))
# Fold sizes should be roughly equal
table(apistrat$.foldID)
#
# one-stage cluster sample
apiclus1 <- cbind(apiclus1,
                  .foldID = folds.svy(apiclus1, nfolds = 5, clusterID = "dnum"))
# For any given cluster, all its observations will be in the same fold;
# and each fold should contain roughly the same number of clusters
with(apiclus1, table(dnum, .foldID))
# But if cluster sizes are unequal,
# the number of individuals per fold will also vary
table(apiclus1$.foldID)
# See the end of `intro` vignette for an example of using such folds
# as part of a custom loop over CV folds
# to tune parameters in a design-consistent random forest model
# }