fillNA: Fill Missing Covariate Values

In secr: Spatially Explicit Capture-Recapture

Fill Missing Covariate Values

Description

Missing covariate values are not allowed in a habitat mask. This function replaces each missing value with a given value or the closest non-missing value. This is not recommended if more than a few cells have missing values.

Usage

fillNA(mask, cov = NULL, method = c("nearest", "value"), value = NULL, verbose = TRUE)

Arguments

mask	habitat mask or similar
cov	character vector of covariate names
method	character
value	character, numeric, or factor value to replace all NA
verbose	logical; if TRUE the replacements are reported

Details

If there is more than one covariate then each is handled separately. If no covariate is named then all covariates are filled.

When method = "value" all NA values are replaced with the given value.

The method argument switches automatically to ‘value’ if value is specified.

Value

An object with the same class as the input and the requested covariate(s) replaced with non-missing values.

See Also

skink, addCovariates

Examples

# artificial example
msk <- make.mask(traps(captdata), spacing = 20)
covariates(msk) <- data.frame(rand = sample(factor(c(NA,1,2)), nrow(msk), 
    replace = TRUE, prob = c(0.2,0.4,0.4)))

msk2 <- fillNA(msk)

par(mfrow = c(1,2))
plot(msk, cov='rand', dots = FALSE)
plot(msk2, cov='rand', dots = FALSE)

msk3 <- fillNA(msk, value = factor(2, levels=1:2))

# a call to fillNA shows whether any covariate has missing values
# assign result to tmp to avoid unwanted output
tmp <- fillNA(msk2)

secr documentation built on Oct. 11, 2026, 5:06 p.m.