problems with coercing a factor to be numeric
On Jan 23, 2013, at 1:58 AM, Francesco Sarracino wrote:
Thanks, this works! but I am surprised that R has such a strange behavior and that there is no way to control it. BTW, also as.integer(pp)-1 works! Still, it doesn't look to me as a first best. At any rate, thanks a lot for your help.
I think it is rather strange that you are criticising R because the mean or sum functions won't coerce factors to numeric class. R is already very loosely typed. It has a fairly limited number of object classes and there is widespread class coercion when it is appropriate. Can you explain why you believed factors or by logical extension character classed variables should get implicitly coerced by all mathematical functions?
David.
> f.
>
>
> On 23 January 2013 10:53, D. Rizopoulos <d.rizopoulos at erasmusmc.nl>
> wrote:
>
>> check also
>>
>> pp <- rep(0:1, 10)
>> pp <- factor(pp, levels=(0:1), labels=c("no","yes"))
>>
>> unclass(pp)
>> unclass(pp) - 1
>>
>>
>> Best,
>> Dimitris
>>
>>
>> On 1/23/2013 10:48 AM, Francesco Sarracino wrote:
>>> Dear Dimitris,
>>>
>>> thanks for your quick reply. I've tried the solutions proposed in
>>> 7.10
>>> How do I convert factors to numeric?
>>>
>>> as.numeric(as.character(pp))
>>> and
>>> as.numeric(levels(pp))[as.integer(pp)]
>>>
>>> However, whatever I do, I get "Warning message: NAs introduced by
>> coercion"
>>> and the output is a vector of NA.
>>>
>>> Any ideas?
>>> f.
>>>
>>>
>>>
>>> On 23 January 2013 10:39, D. Rizopoulos <d.rizopoulos at erasmusmc.nl
>>> <mailto:d.rizopoulos at erasmusmc.nl>> wrote:
>>>
>>> Check R FAQ 7.10: How do I convert factors to numeric?
>>>
>>>
>>> I hope it helps.
>>>
>>> Best,
>>> Dimitris
>>>
>>>
>>> On 1/23/2013 10:33 AM, Francesco Sarracino wrote:
>>>> Dear R listers,
>>>>
>>>> I am trying to compute the mean of a dummy variable that is
>>> encoded as a
>>>> factor. However, even though the levels of my factor are 0 - 1,
>>> when I
>>>> compute the mean (after coercing the factor to be
>>>> numeric), R changes 0 into 1 and 1 into yes, thus altering my
>>> expected
>>>> result.
>>>>
>>>> Please, consider the following working example:
>>>> pp <- rep(0:1, 10)
>>>> pp <- factor(pp, levels=(0:1), labels=c("no","yes"))
>>>> mean(pp) #this won't work because the argument is not numeric or
>>> logical
>>>> mean(as.integer(pp)) # this computes the average, but not on the
>>> range 0-1,
>>>> but 1-2. Indeed, the result is 1.5 and not 0.5 as expected.
>>>>
>>>> What am I doing wrong?
>>>> Thanks in advance for your kind support,
>>>> f.
>>>>
>>>>
>>>
>>> --
>>> Dimitris Rizopoulos
>>> Assistant Professor
>>> Department of Biostatistics
>>> Erasmus University Medical Center
>>>
>>> Address: PO Box 2040, 3000 CA Rotterdam, the Netherlands
>>> Tel: +31/(0)10/7043478 <tel:%2B31%2F%280%2910%2F7043478>
>>> Fax: +31/(0)10/7043014 <tel:%2B31%2F%280%2910%2F7043014>
>>> Web: http://www.erasmusmc.nl/biostatistiek/
>>>
>>>
>>>
>>>
>>> --
>>> Francesco Sarracino, Ph.D.
>>> https://sites.google.com/site/fsarracino/
>>
>> --
>> Dimitris Rizopoulos
>> Assistant Professor
>> Department of Biostatistics
>> Erasmus University Medical Center
>>
>> Address: PO Box 2040, 3000 CA Rotterdam, the Netherlands
>> Tel: +31/(0)10/7043478
>> Fax: +31/(0)10/7043014
>> Web: http://www.erasmusmc.nl/biostatistiek/
>>
>
>
>
> --
> Francesco Sarracino, Ph.D.
> https://sites.google.com/site/fsarracino/
>
> [[alternative HTML version deleted]]
>
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
David Winsemius, MD
Alameda, CA, USA