Skip to content
Prev 275593 / 398506 Next

sometimes removing NAs from code

William Dunlap 
Instead of
   d[-which(condition)]
use
   d[!condition]
where 'condition' is a logical vector.

which(condition) returns integer(0) (an integer vector
of length 0) if there are no TRUEs in 'condition'.
-integer(0) is identical to integer(0) and d[integer(0)]
means to select zero elements from d.

!condition means to flip the senses of all the TRUEs and
FALSEs (and to leave NAs alone) so d[!condition] returns
the elements of d for which condition is not TRUE (along
with NA's for NA's in condition, but you won't have any
of them in your example).

By the way, your use of apply() slows things down and
might lead to errors.  Try replacing
  apply(adata[,1:2],1,function(x)any(is.na(x))))
by
  is.na(adata$y) | is.na(adata$z)
or
  rowSums(is.na(adata[,1:2])) > 0

Bill Dunlap
Spotfire, TIBCO Software
wdunlap tibco.com

Thread (7 messages)

Schatzi sometimes removing NAs from code Oct 26 Natalie Van Zuydam sometimes removing NAs from code Oct 26 Marc Schwartz sometimes removing NAs from code Oct 26 Sarah Goslee sometimes removing NAs from code Oct 26 jim holtman sometimes removing NAs from code Oct 26 William Dunlap sometimes removing NAs from code Oct 26 Schatzi sometimes removing NAs from code Oct 26