Re: How to weight cases in R

On Jun 21, 7:59 am, raquelrgu...@xxxxxxxxx wrote:
On 20 jun, 13:26, John Kane <jrkrid...@xxxxxxxxx> wrote:



You would be better off asking this question on the R-help mailing
list. Seehttps://stat.ethz.ch/mailman/listinfo/r-help

Have a look at ?tranform again. transform is intended to work on a
data.frame and you are feeding it two elements of a data.frame or so
it appears.

The term 'mulheres$pesop' implies that mulheres is a data.frame or
list not a "object" (i.e. vector) within a data.frame.

If mulheres is a data.frame this might work:

logsalhpm <-transform(mulheres,
logsalhp=(mulheres$logsalh)*(mulheres$pesop)/sum(mulheres$pesop))

John Kane, Kingston ON Canada- Ocultar texto entre aspas -

Thanks for your help, John, but unfortunately, it doesn't work...
Here's the message error:

mulheres <-read.table("base81_2_m_r.dat")
logsalhpm <-transform(mulheres, logsalhp=(mulheres$logsalh)*(mulheres$pesop)/sum(mulheres$pesop))

Erro em data.frame(`_data`, e[!matched]) :
argumentos implicam em número de linhas conflitantes: 52012,
0

If the number of lines is conflicting (52012 is the number of
observations), and there's no missing values, what should be the
problem?

Also I have subscribed on the R-list you recommended me. :)

Thanks again, Raquel

I saw the post over in R.

I think you need to construct a small example of the data that you
have and post it. You'll get a couple of comments like this on the R
list (possibly a bit acerbic. :)

To diagnose what is happening is difficult when we don't have any idea
of the structure of the data. As I mentioned earlier it looks like
you have two data.frames (mulares & homens) or maybe a two element
list. If so we need to know what the elements are (matrix, data.frame
etc. You almost certainly don't have a data.frame with two
data.frames inside it.

I'd suggest creating a very small data.frame (perhaps 4 variables and
5 6 subjects and try working with it. It may show you where the error
is, and if you still get the error it will give the people on the R-
help list a lot more to work with.

I had hoped that just using mulheres would work but since it doesn't I
still cannot tell if it is a data.frame or not.

Something I forgot : do a
class(mulheres) and see if you get 'data.frame' as a response.

Also just do a head(mulheres$logsalh) and see what happens.

.