R help - Feb 2011 - Removing -Inf values from all the colums in a dataframe

Hi there,

b is my dataframe.

I have a dataframe. I am trying to get rid of the -INF values but didnt have
much luck

b[which(is.finite(b))]

I can get rid of it for a single column

b[,1][which(is.finite(b[,2]))] but not for all dataframe I used it inside of
a sapply but still no dice

my guess is it might have some differing row numbers and just ignoring the
columns itself. 

Thanks
Ramya




-- 
View this message in context:
http://r.789695.n4.nabble.com/Removing-Inf-values-from-all-the-colums-in-a-dataframe-tp3322059p3322059.html
Sent from the R help mailing list archive at Nabble.com.

On 2011-02-23 19:27, Ramya wrote:
>
> Hi there,
>
> b is my dataframe.
>
> I have a dataframe. I am trying to get rid of the -INF values but didnt
have
> much luck
>
> b[which(is.finite(b))]
>
> I can get rid of it for a single column
>
> b[,1][which(is.finite(b[,2]))] but not for all dataframe I used it inside
of
> a sapply but still no dice
>
> my guess is it might have some differing row numbers and just ignoring the
> columns itself.
>
> Thanks
> Ramya
Note that R does not have "-INF" values.

I'm not sure that I understand what you mean by 'get rid of';
a reproducible example would help.

   b <- data.frame(x = 1:5, y = 11:15)
   b[2, 1] <- b[3, 2] <- -1/0
   b

Here are a couple of ways to delete all cases that
include a -Inf:
1.
   idx <- apply(b, 1, function(x) all(is.finite(x)))
   b[idx, ]

2.
   b[is.finite(rowSums(b)), ]

This assumes that all of your variables are numeric; if not,
you'll have to subset appropriately but, as I said, a reproducible
example would ....

Peter Ehlers