[R] Problems using log() in a plm() regression.

[Milan Bouchet-Valat]

Le dimanche 16 juin 2013 à 09:05 +0200, Guylaine NOUWOUE a écrit :
> Hello,
> 
> I am performing a panel model and I have an error. In fact, I would like to
> set up the pooling model and I have an error message. Any help
> 
> >pool<plm(panel$lexportijt~ldistit+lpopit+lpopjt+lpibit+lpibjt+lpibcit+lpibc
> jt+langueij+histoireij+infrasij+frontiereij+simijt+enclaviit,data=panel,inde
> x=c("origine","annee"))
> 
> Erreur dans parse(text = x) : 
> 
>   fin d'entre inattendu(e) dans "panel$lexportijt ~ ldistit + lpopit +
> lpopjt + lpibit + lpibjt +     lpibcit + lpibcjt + langueij + histoireij +
> infrasij + frontiereij +     simijt + enclaviit +  + "
This is not a problem with log() at all... this is a mere syntax error.

First, there is no need to repeat "panel$" since you specify a data
argument. Second, I think you should add more spacing and line breaks to
your code to make it readable.

Finally, the real bug is that you typed "<" instead of "<-". I think you
should read an introduction to the R language.

This should do what you want :

pool <- plm(lexportijt ~ ldistit + lpopit + lpopjt + lpibit + lpibjt + lpibcit
                       + lpibcjt + langueij + histoireij + infrasij + frontiereij
                       + simijt + enclaviit,
            data=panel, index=c("origine", "annee"))


Regards