[R] add an idx column to the matrix

[Sarah Goslee]

On Wed, Aug 12, 2015 at 2:04 PM, Lida Zeighami <lid.zigh at gmail.com> wrote:
> I applied this code in a loop function but since in some matrices there
> isn't any 2, so I got the below error:
>
> idx<- apply(lofGT_met,1, function(x)as.numeric(any(x==2 & !is.na(x))))
>
>
> Error in apply(lofGT_met, 1, function(x){(any(x == 2)}) :
>   dim(X) must have a positive length

That has nothing to do with whether there are any values of 2 in the
data frame. The code I suggested can handle that:


no2 <- structure(list(X125 = c(0L, 1L, 1L, 0L, 1L), X255 = c(1L, 1L,
 1L, 1L, 0L), X558 = c(0L, 0L, 1L, 0L, 0L), X2366 = c(NA, NA,
 1L, NA, 0L), X177 = c(0L, 0L, 0L, 0L, 0L), X255.1 = c(0L, 1L,
 0L, 0L, 0L)), .Names = c("X125", "X255", "X558", "X2366", "X177",
 "X255.1"), class = "data.frame", row.names = c("aa", "bb", "cs",
 "de", "gh"))

no2
apply(no2, 1, function(x)as.numeric(any(x == 2 & !is.na(x))))


> no2
   X125 X255 X558 X2366 X177 X255.1
aa    0    1    0    NA    0      0
bb    1    1    0    NA    0      1
cs    1    1    1     1    0      0
de    0    1    0    NA    0      0
gh    1    0    0     0    0      0
> apply(no2, 1, function(x)as.numeric(any(x == 2 & !is.na(x))))
aa bb cs de gh
 0  0  0  0  0

It also works for rows that are entirely NA. Thus, I'm forced to
conclude that there's something odd about lofGT_met, and you'll need
to provide more information about that data frame, and a reproducible
example, for a solution to be offered.

Sarah


>
> would you please let me know how to correct it?
> Thanks
>
> On Mon, Aug 10, 2015 at 3:27 PM, Sarah Goslee <sarah.goslee at gmail.com>
> wrote:
>>
>> Easy enough (note that your column names are problematic, though)
>>
>> > mydata <- structure(list(X125 = c(0L, 1L, 2L, 0L, 2L), X255 = c(1L, 1L,
>> + 1L, 1L, 0L), X558 = c(0L, 0L, 2L, 0L, 0L), X2366 = c(NA, NA,
>> + 1L, NA, 0L), X177 = c(0L, 0L, 0L, 0L, 0L), X255.1 = c(0L, 1L,
>> + 0L, 0L, 0L)), .Names = c("X125", "X255", "X558", "X2366", "X177",
>> + "X255.1"), class = "data.frame", row.names = c("aa", "bb", "cs",
>> + "de", "gh"))
>> > mydata$idx <- apply(mydata, 1, function(x)as.numeric(any(x == 2 &
>> > !is.na(x))))
>> > mydata
>>    X125 X255 X558 X2366 X177 X255.1 idx
>> aa    0    1    0    NA    0      0   0
>> bb    1    1    0    NA    0      1   0
>> cs    2    1    2     1      0      0   1
>> de    0    1    0    NA    0      0   0
>> gh    2    0    0     0      0      0   1
>>
>> Sarah
>>
>> On Mon, Aug 10, 2015 at 4:11 PM, Lida Zeighami <lid.zigh at gmail.com> wrote:
>>>
>>> Hi there,
>>>
>>> I have a matrix contain 0,1,2, NA elements.
>>> I want to add a column to this matrix with name of "idx" . then for each
>>> row, I should put 1 in this column (idx) if there is at least one 2 in
>>> that
>>> row otherwise I should put 0 in this column!
>>>
>>> for example  mydata:
>>>
>>>        125   255   558   2366   177    255
>>> aa    0        1       0         NA    0         0
>>> bb    1        1       0         NA    0         1
>>> cs     2        1       2         1       0         0
>>> de    0        1       0         NA    0         0
>>> gh    2       0       0         0        0         0
>>>
>>>
>>> my output should be:
>>>
>>>
>>>        125   255   558   2366   177    255    idx
>>> aa    0        1       0         NA    0         0      0
>>> bb    1        1       0         NA    0         1      0
>>> cs     2        1       2         1       0         0     1
>>> de    0        1       0         NA    0         0      0
>>> gh    2       0       0        2        0         2       1
>>>