[R] Replace NAs in a data frame with elements randomly generated from the beta distribution

[Sibusiso Ndzukuma]

Thank you so much for your assistance. A.K's suggestion is exactly what I was
looking for.

Thanks so much...


On Wed, 10 Apr 2013 23:12:25 +0200 peter dalgaard <pdalgd at gmail.com> wrote

> On Apr 10, 2013, at 13:09 , Sibusiso Ndzukuma wrote:
> 
> > Hi there!
> > 
> > Please help me, I am trying to replace all NAs in a data frame with
numbers
> > randomly generated from the beta(0.1,1) distribution. Firstly I tried
> > looping: 
> > MAT <-
> >
> >
c(0.68,9.86,1.29,0.25,5.28,1.13,1.66,0.41,2.65,0.00,11.5,32.67,0.98,3.06,2.97
> >
,0.51,13.62,1.23,0.00,1.79,5.89,0.00,10.36,1.20,2.06,0.16,0.00,0.00,7.95,3.82
> >
,0.64,0.93,0.73,0.00,5.43,1.45,1.36,1.61,2.02,0.00,10.36,1.51,1.04,1.93,1.40,
> > 0.00,6.03,1.30,1.74,7.07,0.00,0.00,9.20,0.19) 
> > MAT <- data.frame(matrix(MAT, ncol = 6, byrow = FALSE))
> > 
> > MAT[MAT == 0] <- NA
> > 
> > M <- data.frame()
> > 
> > for (i in (1:nrow(MAT)))
> > {
> >     for(j in (1:ncol(MAT)))
> >     {
> >        if(!is.na(MAT))
> >        {
> >           M[i,j] <- MAT[i,j]
> >        }
> >        else
> >           M[i,j] <- rbeta(1,0.1,1)
> >     }
> > 
> > }
> > But this does not work.
> > 
> > Then I tried this:
> > 
> > MAT[is.na(MAT)] <- rbeta(1,0.1,1) 
> > 
> > The problem with the second attempt is that the NAs are replaced with a
> > value which is the same for all NA elements, yet I want them to be
> > different. Please help! 
> > 
> 
> 
> You only asked for one value.... 
> 
> How about
> 
> N <- sum(is.na(MAT))
> MAT[is.na(MAT)] <- rbeta(N, 0.1, 1) 
> 
> -- 
> Peter Dalgaard, Professor,
> Center for Statistics, Copenhagen Business School
> Solbjerg Plads 3, 2000 Frederiksberg, Denmark
> Phone: (+45)38153501
> Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com



____________________________________________________________
South Africas premier free email service - www.webmail.co.za 

Great Renault Deals! http://www.interfacebusinesses.co.za/20054319/