Perhaps I'm missing something obvious.
In R, TRUE && NA evalues to NA. This doesn't make sense to me, because && should evaluate left to right, and stop as soon as one of its conditions is true.
Asked
Active
Viewed 3,741 times
4
flodel
- 87,577
- 21
- 185
- 223
Daniel Kessler
- 1,172
- 17
- 21
-
It's touched on in [this answer](http://stackoverflow.com/a/1535492/288545) but I still don't understand *why* – Daniel Kessler May 30 '13 at 19:32
-
see http://stackoverflow.com/questions/16844139/na-true-and-false-properties and especially @joran answer about ambiguity of `NA`. – Beasterfield May 30 '13 at 19:33
-
6AND requires both conditions to be true, so will only short-circuit on `FALSE` – James May 30 '13 at 19:39
-
http://spikedmath.com/445.html , http://mrburkemath.blogspot.ca/2011/05/coffee-logic.html – Ben Bolker May 30 '13 at 20:17
1 Answers
21
This doesn't make sense to me, because && should evaluate left to right, and stop as soon as one of its conditions is true.
This is wrong. You are mixing up && with ||:
TRUE && FALSEgivesFALSE- && requires both conditions to be TRUE
&&will short-circuit on FALSE
TRUE || FALSEgivesTRUE||requires a single condition to be TRUE||will short-circuit on TRUE
Also,
TRUE || NA
gives
TRUE
csgillespie
- 59,189
- 14
- 150
- 185