[R] does package "QuantPsych" function lm.beta can handle resultsof a regression with weights?

Tom Fletcher tom.fletcher.mp7e at statefarm.com
Mon Jul 26 15:24:37 CEST 2010


The original function was created for a simple example. It never was
written to address weighted regression. A quick fix will work for you
situation.

### The original is:


lm.beta <-
function (MOD) 
{
    b <- summary(MOD)$coef[-1, 1]
    sx <- sd(MOD$model[-1])
    sy <- sd(MOD$model[1])
    beta <- b * sx/sy
    return(beta)
}

####  A newer modification:

lm.betaW <- 
function (MOD) 
{
    b <- summary(MOD)$coef[-1, 1]
    sx <- sd(MOD$model[-1][1])
    sy <- sd(MOD$model[1])
    beta <- b * sx/sy
    return(beta)
}

The above should do the trick.

TF

-----Original Message-----
From: r-help-bounces at r-project.org [mailto:r-help-bounces at r-project.org]
On Behalf Of Dimitri Liakhovitski
Sent: Thursday, July 22, 2010 12:35 PM
To: r-help at r-project.org
Subject: [R] does package "QuantPsych" function lm.beta can handle
resultsof a regression with weights?

Hello, and sorry for not providing an example.
I run a regular linear regression (using lm) and use weights with it
(weights = ...).
I use "QuantPsych" package, its function lm.beta to extract standardized
regression weights from my lm regression object.

When I don't use weights, everything is fine.
But when I do use weights, I get an error that refers to lm.beta code:
"In b * sx : longer object length is not a multiple of shorter object
length"

This happens because there is an extra column in the object:
regr$model that lm.beta is using to get at the betas.
Is there some other package that just gives me the standardized
regression weights - even if I used weights for regression?

Thank you!

--
Dimitri Liakhovitski
Ninah Consulting
www.ninah.com

______________________________________________
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.



More information about the R-help mailing list