unzero: Remove precipitation zeros by setting them to NA

In sethmcg/climod: Bias correction and other tools for climate model output

Description

unzero eliminates zero precipitation data from the bias correction by setting zero values to NA, which is ignored. rezero sets NA values back to zero after bias correction.

Usage

unzero(x, warn=FALSE)

rezero(x, refloor=TRUE)

Arguments

x	A vector of numeric values
warn	Logical; warn if NA values already present in input
refloor	Logical; also set negative values to zero

Details

The zeros in precipitation data cause problems for distribution mapping. Changing zeros to NA allows them to pass harmlessly through the rest of the bias correction procedure without requiring a lot of complicated and error-prone indexing to remove and reinsert them from the arrays.

It is not uncommon for precipitation data to have small negative values after bias correction. By default rezero also sets those values to zero when returning NA values back to zero.

Examples

set.seed(22)

x <- rbinom(10, 2, 0.2)
y <- unzero(x)
x
y
rezero(y)
z <- unzero(c(x, NA, NA), warn=TRUE)
z <- c(z, -1)
z
rezero(z, refloor=FALSE)

sethmcg/climod documentation built on Nov. 19, 2021, 11:12 p.m.