[R-sig-ME] Factor conversion
ONKELINX, Thierry
Thierry.ONKELINX at inbo.be
Thu Jun 9 11:12:53 CEST 2011
Dear Iker,
It looks like a problem with the txt file to me. You are expecting numerical values but you get characters instead. Use str() and summary() to check if your dataset contains what it should contain. If not, you need to find out what is going wrong. read.table choses the class of the variable depending on the data is finds for that variable. The least complex data class is used. But when you have one non-numerical data entry in a vector of numbers, then the whole vector will be converted to a character vector or factor.
So the problem is most likely not with the software but with your datafile.
Best regards,
Thierry
----------------------------------------------------------------------------
ir. Thierry Onkelinx
Instituut voor natuur- en bosonderzoek
team Biometrie & Kwaliteitszorg
Gaverstraat 4
9500 Geraardsbergen
Belgium
Research Institute for Nature and Forest
team Biometrics & Quality Assurance
Gaverstraat 4
9500 Geraardsbergen
Belgium
tel. + 32 54/436 185
Thierry.Onkelinx at inbo.be
www.inbo.be
To call in the statistician after the experiment is done may be no more than asking him to perform a post-mortem examination: he may be able to say what the experiment died of.
~ Sir Ronald Aylmer Fisher
The plural of anecdote is not data.
~ Roger Brinner
The combination of some data and an aching desire for an answer does not ensure that a reasonable answer can be extracted from a given body of data.
~ John Tukey
> -----Oorspronkelijk bericht-----
> Van: r-sig-mixed-models-bounces at r-project.org [mailto:r-sig-mixed-models-
> bounces at r-project.org] Namens Iker Vaquero Alba
> Verzonden: woensdag 8 juni 2011 0:14
> Aan: r-sig-mixed-models at r-project.org
> Onderwerp: [R-sig-ME] Factor conversion
>
>
>
> Dear R list:
>
> I am fitting a model where I have several covariates which are coefficients of
> variation. For some reason, the software interprets them as factors. I was
> converting them with the function "as.numeric", but when doing so, the variable
> is converted into an ordinal categorical one.
> I have tried the “as.is” function but I get this message:
>
> In read.table("Coefvar insects 2009+2010.txt", header = T, as.is = "bithrcoef")
> :
> not all columns named in 'as.is' exist
>
> Could you tell me wich is the most suitable converting function I should use?
> I’ve tried “as.vector” and “as.character”, but I get this error messages when
> fitting the model with lme or lmer:
> With lmer:
>
> “insectmodel2<-
> lmer(inslarge~tlengthcoef+brivencoef+brithrcoef+condcoef+sex+tlengthcoef:se
> x+brivencoef:sex+brithrcoef:sex+condcoef:sex+tlengthcoef:brivencoef+tlengthc
> oef:brithrcoef+tlengthcoef:condcoef+brivencoef:brithrcoef+brivencoef:condco
> ef+brithrcoef:condcoef+(1|year))”
>
> Error in mer_finalize(ans) : Downdated X'X is not positive definite, 35.
> Also: lost warning messages
> 1: In model.matrix.default(mt, mf, contrasts) :
> variable 'tlengthcoef' converted to a factor
> 2: In model.matrix.default(mt, mf, contrasts) :
> variable 'brivencoef' converted to a factor
> 3: In model.matrix.default(mt, mf, contrasts) :
> variable 'brithrcoef' converted to a factor
> 4: In model.matrix.default(mt, mf, contrasts) :
> variable 'condcoef' converted to a factor
>
> With lme:
>
> “insectmodel2<-
> lme(inslarge~tlengthcoef+brivencoef+brithrcoef+condcoef+sex+tlengthcoef:sex
> +brivencoef:sex+brithrcoef:sex+condcoef:sex+tlengthcoef:brivencoef+tlengthc
> oef:brithrcoef+tlengthcoef:condcoef+brivencoef:brithrcoef+brivencoef:condco
> ef+brithrcoef:condcoef,random=~1|year)”
>
> Error in MEEM(object, conLin, control$niterEM) :
> Singularity in backsolve at level 0, block 1
> Also: Lost warning messages
> 1: In model.matrix.default(fixed, data = X) :
> variable 'tlengthcoef' converted to a factor
> 2: In model.matrix.default(fixed, data = X) :
> variable 'brivencoef' converted to a factor
> 3: In model.matrix.default(fixed, data = X) :
> variable 'brithrcoef' converted to a factor
> 4: In model.matrix.default(fixed, data = X) :
> variable 'condcoef' converted to a factor
>
>
> Any help will be greatly appreciated.
>
> Thank you very much. Iker
> [[alternative HTML version deleted]]
More information about the R-sig-mixed-models
mailing list