Re: [R-sig-phylo] compar.gee: how to extract SE, t, and P from output?

[Nina Hobbhahn]

Hi Emmanuel,

thanks for your reply. Works mostly, however, I get an error message I can't 
seem to figure out:

"Error in coef[, 2] <- sqrt(diag(x$W)) : replacement has length zero."

What does R mean or want here?

Thanks again,

Nina



On 2012-05-09, at 3:19 PM, Emmanuel Paradis wrote:

> Hi Nina,
> 
> You cannot do it directly: you have to get the code from print.compar.gee; 
> from R:
> 
> > print.compar.gee
> 
> From this you can create a function, eg:
> 
> foo <- function(x)
> {
>    nas <- is.na(x$coef)
>    coef <- x$coef[!nas]
>    cnames <- names(coef)
>    coef <- matrix(rep(coef, 4), ncol = 4)
>    dimnames(coef) <- list(cnames,
>                           c("Estimate", "S.E.", "t", "Pr(T > |t|)"))
>    df <- x$dfP - dim(coef)[1]
>    coef[, 2] <- sqrt(diag(x$W))
>    coef[, 3] <- coef[, 1]/coef[, 2]
>    coef
> }
> 
> Then do:
> 
> foo(GEE)
> 
> HTH.
> 
> Best,
> 
> Emmanuel
> 
> Nina Hobbhahn wrote on 08/05/2012 18:55:
>> Dear fellow list users,
>> 
>> I would like to extract the values for SE, t, and Pr from the compar.gee 
>> output, but seem to unable to do so. The command lines
>> 
>> GEE<-­‐compar.gee(trait~reward-­‐1, phy=phy2, data=DF.Disa, 
>> family=gaussian)
>> 
>> GEE
>> 
>> return the following output:
>> 
>> Beginning Cgee S-­‐function, @(#) geeformula.q 4.13 98/01/27 running glm 
>> to get initial regression estimate
>> 
>> rewardN rewardY
>> 2.689142 3.421509
>> Beginning Cgee S-­‐function, @(#) geeformula.q 4.13 98/01/27 running glm 
>> to get initial regression estimate
>> 
>> rewardN rewardY
>> 2.689142 3.421509
>> Call: compar.gee(formula = trait ~ reward -­‐ 1, data = DF.Disa,
>> 
>> family = gaussian, phy = phy2) Number of observations: 31
>> 
>> Model:
>> Link: identity
>> 
>> Variance to Mean Relation: gaussian
>> 
>> 
>> QIC: 165.2011
>> 
>> Summary of Residuals:
>> Min 1Q Median 3Q Max
>> 
>> 3.1462585 0.2560335 1.5256031 2.2012787 4.6689764
>> 
>> Coefficients:
>> Estimate S.E. t Pr(T>  |t|)
>> 
>> rewardN 1.959375 1.091801 1.794626 0.1016430 rewardY 1.908384 1.117150 
>> 1.708261 0.1170666
>> 
>> Estimated Scale Parameter: 4.862036 "Phylogenetic" df (dfP): 12.45332
>> 
>> When typing
>> 
>> GEE$coefficients
>> 
>> I get
>> 
>> 
>> rewardN rewardY
>> 
>> 1.959375 1.908384
>> 
>> 
>> but I'm unable to access SE, t, and Pr(T>  |t|). Please can any of you help? 
>> Thank you very much,
>> 
>> 
>> Nina
>> 
>> 
>> 
>> 
>> Dr. Nina Hobbhahn
>> Post-doctoral fellow
>> Lab of Prof. S. D. Johnson
>> School of Life Sciences
>> University of KwaZulu-Natal
>> Private Bag X01
>> Scottsville, Pietermaritzburg, 3201
>> South Africa
>>      [[alternative HTML version deleted]]
>> 
>> 
>> 
>> 
>> _______________________________________________
>> R-sig-phylo mailing list
>> R-sig-phylo@r-project.org
>> https://stat.ethz.ch/mailman/listinfo/r-sig-phylo
> 
> -- 
> Emmanuel Paradis
> IRD, Jakarta, Indonesia
> http://ape.mpl.ird.fr/

        [[alternative HTML version deleted]]


_______________________________________________
R-sig-phylo mailing list
R-sig-phylo@r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-phylo