[Rd] columnames changes behaviour of formula

[Joshua Wiley]

P.S. It really is sloppy code to mix variables from the global
environment with those inside a data frame. I.e.:

coef(lm(d ~ -1 + (.)^2, data = x))

the only time I think it makes sense to have different objects for the
outcome and predictors are when for speed purposes, you are using a
low level function, such as lm.fit or fastLmPure from the RcppEigen
package.


On Thu, May 24, 2012 at 9:46 PM, Joshua Wiley <jwiley.psych at gmail.com> wrote:
> Hi Robin,
>
> Seems like the intended behavior to me.  From the docs:
> "There are two special interpretations of '.' in a formula.  The usual
> one is in the context of a 'data' argument of model fitting functions
> and means 'all columns not otherwise in the formula' "
>
> d is in the formula so the only column not in the formula is nd.  the
> (.)^2 asks for all two way interactions, but with only one variable,
> there are none.
>
> What were you expecting?
>
> Josh
>
> On Thu, May 24, 2012 at 9:25 PM, robin hankin <hankin.robin at gmail.com> wrote:
>> Hello. precompiled R-2.15.0, svn58871, macosx 10.7.4.
>>
>>
>> I have discovered that defining column names of a dataframe can alter the
>> behaviour of lm():
>>
>>
>> d <- c(4,7,6,4)
>> x <- data.frame(cbind(0:3,5:2))
>> coef(lm(d~ -1 + (.)^2,data=x))
>>   X1    X2 X1:X2
>> -1.77  0.83  1.25
>> R>
>> R>
>>
>>
>> OK, so far so good.  But change the column names of 'x' and the behaviour
>> changes:
>>
>>
>> colnames(x) <- c("d","nd")   # 'd' == 'death' and 'nd' == 'no death'
>> coef(lm(d~ -1 + (.)^2,data=x))
>>       nd
>> 0.2962963
>>
>>
>>
>> I am not sure if this is consistent with the special meaning of '.'
>> described under ?formula.
>>
>> Is this the intended behaviour?
>>
>>
>> --
>> Robin Hankin
>> Uncertainty Analyst
>> hankin.robin at gmail.com
>>
>>        [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel
>
>
>
> --
> Joshua Wiley
> Ph.D. Student, Health Psychology
> Programmer Analyst II, Statistical Consulting Group
> University of California, Los Angeles
> https://joshuawiley.com/



-- 
Joshua Wiley
Ph.D. Student, Health Psychology
Programmer Analyst II, Statistical Consulting Group
University of California, Los Angeles
https://joshuawiley.com/