Odp: NA-values and logical operation

Hi Petr,

works perfect, thanks a lot!

Perhaps you can take a look at this 'problem'? I want to get a
new vector for which (x == 2 & y == 2 & z == 2) = TRUE
(the real dataframe has 21 columns to check)

o <- data.frame(x=c(2,2,2,2,2), y=c(1,2,2,2,NA), z=c(2,NA,2,1,2))

#> o == 2
#        x     y     z
#[1,] TRUE FALSE  TRUE
#[2,] TRUE  TRUE    NA
#[3,] TRUE  TRUE  TRUE
#[4,] TRUE  TRUE FALSE
#[5,] TRUE    NA  TRUE

apply(o == 2 , 1, function(x) isTRUE(x))
# [1] FALSE FALSE FALSE FALSE FALSE

At least I would expect the third index to be TRUE...

apply(o == 2 , 1, function(x) all(x))

rowSums(o==2)

rowSums(o==2, na.rm=T)

Best
  Patrick


Zitat von Petr PIKAL <petr.pikal at precheza.cz>:

Hi


r-help-bounces at r-project.org napsal dne 13.01.2009 11:58:45:

Dear list,

as a result of a logical operation I want to assign
a new variable to a DF with NA-values.

z <- data.frame( x = c(5,6,5,NA,7,5,4,NA),
                  y = c(1,2,2,2,2,2,2,2) )

p <- (z$x <= 5) & (z$y == 1)
p
z[p, "p1"] <-5
z
# ok, this works fine

z <- z[,-3]

p <- (z$x <= 5) & (z$y == 2)
p
z[p, "p2"] <-5
z
# this failed... - how can I assign the value '5' to the new
# var "p2"

z[which(p), "p2"] <-5

works.

Regards
Petr

Thanks for any help!!
Patrick