[R] using string as variable name in model

[Roni Kobrosly]

update() works!
Thank you thomas


On Jun 4, 2010, at 11:47 AM, Thomas Lumley wrote:

> 
> 
> I think you're looking for the update() function.
> 
>   -thomas
> 
> On Thu, 3 Jun 2010, Roni Kobrosly wrote:
> 
>> Hi,
>> 
>> I made a small table of strings that will serve as variable names for lm models I will run. The table looks like this:
>> 
>>> varnames
>>  num                        name
>> 1    1      zCANTAB_log_IED_totaltrials
>> 2    2       zCANTAB_log_IED_preED
>> 3    3        zCANTAB_logPALerrors
>> 4    4           zCANTAB_PALstages
>> 5    5              zCANTAB_logRTI
>> 6    6       zCANTAB_RVP_Totalmisses
>> 7    7      zCANTAB_log_RVP_falsealarm
>> 8    8           zCANTAB_DMS_12000
>> 9    9         zCANTAB_PRM_Percent
>> 10  10         zCANTAB_SRM_Percent
>> 11  11     zCANTAB_sqrt_SWM_within
>> 12  12      zCANTAB_sqrt_SWM_Total
>> 13  13                  zSS_WJ_PC
>> 14  14                 zSS_WJ_CALC
>> 15  15                   zSS_WJ_LW
>> 16  16                   zSS_WJ_AP
>> 17  17                   zSS_WJ_MF
>> 
>> 
>> What I'm trying to do is
>> 
>> 
>> varnames[1,2] -> outcome
>> 
>> lm(outcome ~ income + covariate1 + coviarate2, data=my.data) -> model
>> 
>> I get the following error message:
>> "Error in model.frame.default(formula = outcome ~ Hollings_Enroll + Child_sex +  :
>> variable lengths differ (found for 'Hollings_Enroll')"
>> 
>> But when I run the model:
>> lm(zCANTAB_log_IED_totaltrials ~ income + covariate1 + coviarate2, data=my.data) -> model
>> 
>> it works fine. Eventually, once I resolve this, I'd like to make a custom R function to automatically generate 17 regression models with these 17 outcome variable strings. Do you have any ideas how to resolve this error?
>> 
>> Thanks!
>> -Roni
>> 
>> ______________________________________________
>> 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.
>> 
> 
> Thomas Lumley			Assoc. Professor, Biostatistics
> tlumley at u.washington.edu	University of Washington, Seattle
>