Taking advantage of existing packages

Discuss the jamovi platform, possible improvements, etc.

by Nem2k » Sat Apr 07, 2018 5:57 am

Just heard about this project and it seems really interesting. I'm wondering how easy it would be to take advantage of existing packages that are not implemented in the jmv R package?

Obviously we cant expect jamovi to be able to do everything, but often times we have to do more complicated analyses like mixed-effects models or testing mediation etc. There are already great packages to handle these things in the R ecosystem (e.g. lme4)

Is it possible, now or in the future, for jamovi to conduct statistics using those external packages? I would love a visual interface to those other packages, but as it stands now it seems those more complicated models would first need to be implemented into the jmv package

Possible/more efficient to build a bridge to existing packages, instead of re-implementing things (that other packages already do well) into jmv?
Nem2k
 
Posts: 3
Joined: Sat Apr 07, 2018 5:51 am

by jonathon » Sat Apr 07, 2018 6:43 am

hi,

take a look at the jamovi library:

https://www.jamovi.org/library.html

there's modules for median/moderation, linear mixed effects models, etc.

if you have some skills in R, you can check out the developer hub which describes how to create new modules from existing packages: https://dev.jamovi.org

cheers

jonathon
User avatar
jonathon
 
Posts: 456
Joined: Fri Jan 27, 2017 10:04 am

by deepankardatta » Wed Apr 11, 2018 5:05 pm

In addition to Jonathon (and as not a Jamovi developer) you can add turn any R function into a Jamovi analysis. However someone needs to code the Jamovi analysis for this. As some trying to create additional analysis for Jamovi, it seems to be easier writing the Jamovi module into the actual CRAN R function itself. But Jonathon can advise better!
deepankardatta
 
Posts: 20
Joined: Mon Jan 29, 2018 4:30 am


Return to General