[datatable-help] hetcor error when variables are more than 30

Arunkumar Srinivasan aragorn168b at gmail.com
Wed Nov 25 19:00:56 CET 2015


Hi, you’re on the wrong mailing list. This is the mailing list for R’s data.table package.

-- 
Arun

On 25 November 2015 at 18:23:32, mahadev (orlovoleh at gmail.com) wrote:

Hello friends,  

I'm trying to do a factor analysis of a questionnaire of 243 items with  
YES/NO (1/0) type answers. That's why I need a polychoric correlation  
matrix. I'm trying to use hetcor() function from "polychor" package:  

e <- read.xlsx2(file="ukon_1.xlsx", sheetIndex = 1, as.data.frame="true")  
#this is to load my binary data  
e1 <- hetcor(e) #this is to generate correlation matrix  

And I get an error:  

Error in optim(0, f, control = control, hessian = TRUE, method = "BFGS") :  
non-finite finite-difference value [1]  

When I cut my data to up to 30 variables, everything is fine - I get a  
correlation matrix. Any ideas how to fix that?  

PS Sorry for my English - it's not my native language =(  



--  
View this message in context: http://r.789695.n4.nabble.com/hetcor-error-when-variables-are-more-than-30-tp4715010.html  
Sent from the datatable-help mailing list archive at Nabble.com.  
_______________________________________________  
datatable-help mailing list  
datatable-help at lists.r-forge.r-project.org  
https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/datatable-help  
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.r-forge.r-project.org/pipermail/datatable-help/attachments/20151125/f2086640/attachment.html>


More information about the datatable-help mailing list