Howdy, Stranger!

It looks like you're new here. If you want to get involved, click one of these buttons!

Supported by

Network analysis with Spearman correlations

edited May 13 in JASP & BayesFactor

I am using the latest version 0.19.3, and I need to estimate a network with Spearman correlations using EBICglasso. From the information available, the listed correlation methods are Auto, Cor, Cov, Npn and Spearman. However, in my analysis I do not have the Spearman option (picture below). Is there a bug? How can I do the analysis with Spearman correlations?

Additional question: From what I can tell, the network analysis currently does not have the "R" button, i.e. it is not possible to view the R code directly. Is there another way?

Comments

  • I am commenting just to "bump" this a bit, as I see that other posts get responses - is there a way to do EBICglasso with Spearman correlations in the current version? Or is that something for the future? Hopefully the authors can respond, but maybe someone else had a similar problem and has an idea...

  • I will forward this to the experts. If we mention it in the helpfile and it is missing it might be an oversight and you could issue a bug report on our GitHub page (https://jasp-stats.org/feature-requests-bug-reports/)

    EJ

  • edited May 31

    Thanks for the advice, I submitted the bug report. Could you also address the other issue, is it possible to view the R code for network analysis? I can also submit it as a feature request, but maybe there is another way to peek "under the hood" of JASP when there is no explicit "R" button?

  • There is --all code is open-source-- but when the R button is absent is will be similar to an archeological expedition on GitHub. I recommend that you issue a feature request asking for all code to have the R button.

Sign In or Register to comment.