Added new function condReplace.R

[bodirsky]

Currently, the code often includes code like

x[is.na(x)] <- y[is.na(x)]
which is problematic if x any y have a different dimensionality or just a different ordering of the sets or set elements (even if the sets and set elements are identical). This can lead to serious bugs.
The new function replaces this with a safe function, which also integrates conditional statements with lower dimensionality (e.g. conditions over all years or conditions over all regions), as well as replacement with different dimensionality (e.g. partially replacing a country-level dataset with global values).

[codeZeilen]

The following removes the redundant applications of the condition function and the conversions to logical. Also pulls out the actual replacement.

Suggested change

	condReplace <- function(x, condition, replace) {
	if (is.function(condition)) {
	replace <- magpie_expand(x = as.magpie(replace), ref = x)
	x[condition(x)] <- replace[condition(x)]
	} else {
	replace <- magpie_expand(x = as.magpie(replace), ref = x)
	condition <- magpie_expand(x = as.magpie(condition), ref = x)
	x[as.logical(condition)] <- replace[as.logical(condition)]
	}
	return(x)
	}
	condReplace <- function(x, condition, replace) {
	if (is.function(condition)) {
	replace <- magpie_expand(x = as.magpie(replace), ref = x)
	condition <- as.logical(condition(x))
	} else {
	replace <- magpie_expand(x = as.magpie(replace), ref = x)
	condition <- as.logical(magpie_expand(x = as.magpie(condition), ref = x))
	}
	x[condition] <- replace[condition]
	return(x)
	}