jYS module into Jamovi library

[yurismol]

Dear Jamoviers!
Let me introduce a new jYS module in the Jamovi library (correlation clustering, ROC classification).
I am ready to listen comments and make corrections.

[jonathon]

great work yuri!

jonathon

[mcfanda@gmail.com]

Great!

[Gabriel Tiraboschi]

Thank for this update. I came across this module and I am very impressed!

Can someone kindly clarify what tests are being performed under the hood in the missing values estimation (MAR/MCAR)? I cannot find any documentation on that and it doesn't seem to be Little's test. Thanks!

[yurismol]

This is a translation of the R module missr in jamovi
Little’s missing completely at random (MCAR) test
https://cran.r-project.org/web/packages/missr/missr.pdf

[Gabriel Tiraboschi]

This is a translation of the R module missr in jamovi

Thank you so much for your reply Yuri!

If you allow me one more question: I used this package in R-studio and it doesn't ask me for a "training variable". Could you please clarify why we need a training variable for the Little's test in your module?

Thanks
Gabriel

[yurismol]

Thank you so much for your reply Yuri!

If you allow me one more question: I used this package in R-studio and it doesn't ask me for a "training variable". Could you please clarify why we need a training variable for the Little's test in your module?

Thanks
Gabriel

Training Variables and Incomplete Variables as one dataframe participate in the process of estimation and imputing missing values, but are not displayed in the resulting tables and graphs.
For example, you do not need to generate an imputation variable from a variable without missing values, but you need it for imputation others.
Although this probably introduces some confusion. Maybe I should have called them "Complete Variables"?

[Gabriel Tiraboschi]

Thank you so much Yuri! Complete variables sounds better. If you want to add more clarity to your module, I would suggest to write some instructions/documentation and show it as the user loads the module. The module jReshape is a good example.

Thank you for your work

[yurismol]

Once more minor note:
Jamovi <=2.6.44 sometimes incorrectly calculate = 0 OOB (out-of-bag error) for missRangher method. I didn't find why.
Version 2.7.0 (near future release with R 4.5.0) do it well.

[yurismol]

Thank you so much Yuri! Complete variables sounds better. If you want to add more clarity to your module, I would suggest to write some instructions/documentation and show it as the user loads the module. The module jReshape is a good example.

Thank you for your work

Thanks! It will be done.