R to ignore NULL values - r

Question

I have 2 vector in R, but some of the values in both are marked as "NULL".
I want R to ignore "NULLS", but still "acknowledge" their presence because of indexes ( I´m using intersect and which function).
I have tried this:
for i in 1:length(vector)
if vector=="NULL"
i=i+1
else
'rest of the code'
Is this a good approach? The algorithm is running, but vector are very large.

Answer 1

You should change "NULL" for NA, which is R's native representation for NULL values. Then many functions have ways of dealing with NA values, such as na.action option... You shouldn't call your vector 'vector' since this is a reserved word for the class.
yourvector[yourvector == "NULL"] <- NA
Also you shouldn't add 1 to i in your if, just do nothing:
for (i in 1:length(yourvector)) {
if (!is.na(yourvector[i])) {
#rest of the code
}
}
Also tell what you wanna do. You probably don't need a for.

Answer 2

This code contains several errors:
First off, a vector cannot normally contain NULL values at all. Are you maybe using a list?
if vector=="NULL"
you probably mean if (vector[i] == "NULL"). Even so, that’s wrong. You cannot filter for NULL by comparing to the character string "NULL" – those two are fundamentally different things. You need to use the function is.null instead. Or, if you’re working with an actual vector which contains NA values (not NULL, like I said, that’s not possible), something like is.na.
i=i+1
This code makes no sense – leaving it out won’t change the result because the loop is in charge of incrementing i.
Finally, don’t iterate over indices – for (i in 1 : length(x)) is bad style in R. Instead, iterate over the elements directly:
for (x in vector) {
if (! is.na(x)) {
Perform action
}
}
But even this isn’t very R like. Instead, you would do two things:
use subsetting to get rid of NA values:
vector[! is.na(vector)]
Use one of the *apply functions (for instance, sapply) instead of a loop, and put the loop body into a function:
sapply(vector[! is.na(vector)], function (x) Do something with x)