Re: [R] Pseudo R squared for quantile regression with replicates

2014-09-19 Thread Donia Smaali Bouhlila
Thank you for considering my email and replying. Well, I am working with TIMSS 2007 survey data ( Trends in international mathematics and science study). TIMSS is a stratified DATA, where the primary sampling unit is the school I want to test the effect of some independent variables reflectin

Re: [R] Pseudo R squared for quantile regression with replicates

2014-09-19 Thread Donia Smaali Bouhlila
Thank you for considering my email and replying. Well, I am working with TIMSS 2007 survey data ( Trends in international mathematics and science study). TIMSS is a stratified DATA, where the primary sampling unit is the school I want to test the effect of some independent variables reflectin

Re: [R] Pseudo R squared for quantile regression with replicates

2014-09-18 Thread Anthony Damico
here is a reproducible example, mostly from ?withReplicates. i think something would have to be done using return.replicates=TRUE to manually compute survey-adjusted residuals, but i'm not really sure what nor whether the pseudo r^2 would be meaningful :/ library(survey) library(quantreg) data

Re: [R] Pseudo R squared for quantile regression with replicates

2014-09-18 Thread David L Carlson
It is hard to say because we do not have enough information. R has approximately 6,000 packages and you have not told us which ones you are using. You have not told us much about your data and you have not told us where to find the query from August 2006. The basic problem is that your "fit" is