[R] extracting the t-statistic: just the numbers, please

[Peter Dalgaard]

Michael Rennie <mrennie at utm.utoronto.ca> writes:

> > simt
>          t          t          t          t          t          t          t 
>  0.3474150  0.1542973 -0.4044992  1.2466663 -0.2933944 -0.5809257  0.7799080 
>          t          t          t 
> -1.4132713  1.2048335 -0.6596936
> 
> Which gives me a list, but not in a form that I can do anything with. This is 
> in stark contrast to what happens when requesting p-values, which gives output 
> like this:

..

> > simt
>  [1] 0.6763749 0.7480091 0.9447851 0.3342029 0.7852635 0.3199006 0.5272153
>  [8] 0.3863616 0.7333693 0.7268907
> 
> Now THAT'S what I'd like to get for my t-statistics- a nice vector (simt) that 
> I can deal with later, rather than the output I am currently getting (the first 
> output above).
> 
> Does anyone know a way to extract JUST the t-statistics from the t.test, 
> without the "t" character header, so I can generate a nice little vector? 
> Alternatively, can I manipulate the output I am currently getting for the t-
> statistics so that I can isolate just the numbers? 

It *is* a nice little vector! It's just that it has names. 

names(simt) <- NULL 

gets rid of them. 

BTW, simt <- c(simt, sims) is going to kill performance with large repeat
counts since it copies all previous results every time around.
Preallocate the result, or consider using replicate() or sapply().

-- 
   O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)             FAX: (+45) 35327907