[R] turn regression coefficients into matrix or...

[Norbert NEUWIRTH]

jun,

im am also quite new to R. so i think, this is a question all we R-newbees  
ask  ;-). having had the same "problem" the other day, i "solved" it the  
following way:

####  Multiple Regressions - Tables

data(anscombe)                                      # load anscombe  
dataset (implemented somwhere in R)
x5 <- rnorm(11,14,1)                                # create an additional  
variable (at random)
anscombe.nn <- cbind(anscombe,x5)                   # attach the variable  
to the dataset
attach(anscombe.nn)                                 # attach the dataset  
to the searchpath (just for convenience)
anscombe.nn                                         # have a look on the  
data

ans.reg <- vector(4, mode = "list")                 # create empty list  
(just  for speeding up)

for(i in 1:4){                                      # now the 4  
regressions are stored to the list
     j <- i+1
     x1 <- get(paste("x", i, sep = ""))              # exogenous var. #1
     x2 <- get(paste("x", j, sep = ""))              # exogenous var. #2
     y <- get(paste("y", i, sep = ""))               # endogenous
     ans.reg[[i]] <- glm(y ~ x1+x2,family=gaussian)  # do the regression  
(out of 4)
     print(summary(ans.reg[[i]],cor=FALSE))
}

detach(anscombe)                                    # detach dataset from  
search path
lapply(ans.reg, coef)                               # see each regression  
in one line
sapply(ans.reg, coef)                               # have the  
coefficients in a table
x<-as.matrix(sapply(ans.reg, coef))                 # convert table to  
matrix
x

this solution is quite comparable to Ligges(2007) [published in german and  
japanese, i think]

i have an additional question to list-members, that have left the  
newbee-status yet: how can i get R to hand over the standard errors,  
significance levels etc, so that i can create a table with  coeff and SE  
and sig.?

norbert


Am 13.03.2007, 02:58 Uhr, schrieb Jun Xu <mystata at hotmail.com>:

> I don't have much experience with r. What I am trying to do is to turn
> regression coefficients (after I run a lm or glm model) into some matrix
> such that I can do some post-estimation calculation, for example  
> predicted
> probabilities in glm model, etc.. Or, is there any function in r that I  
> can
> use to do something along that line? thanks.
>
> Jun Xu, Ph.D.
> Department of Sociology
> Ball State University
>
> _________________________________________________________________
> Mortgage rates as low as 4.625% - Refinance $150,000 loan for $579 a  
> month.
> Intro*Terms
> https://www2.nextag.com/goto.jsp?product=100000035&url=%2fst.jsp&tm=y&search=mortgage_text_links_88_h27f6&disc=y&vers=743&s=4056&p=5117
>
> ______________________________________________
> R-help at stat.math.ethz.ch 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.
>



-- 
-------------------------------
Mag. Norbert NEUWIRTH

Roubiczekgasse 2/23
A-1100  WIEN
mob: +43 699 1835 0704