search

ocbe-uio / BayesMallows

R-package for Bayesian preference learning with the Mallows rank model.
https://ocbe-uio.github.io/BayesMallows/
GNU General Public License v3.0
21 stars 9 forks source link

Validate implementation for partial rankings #378

Closed osorensen closed 4 months ago

osorensen commented 6 months ago

Based on discussions with Marta. Reproduce results in JMLR paper. Data are here.

library(RankAggreg) 
data(geneLists)
osorensen commented 6 months ago

The JMLR paper says this:

image

I find it hard to believe we actually used L=40.

crispinomarta commented 6 months ago

This is weird.... I do not remember who wrote that section and run the final codes: maybe Valeria must be involved in this investigation?

Il giorno ven 16 feb 2024 alle ore 12:05 Øystein Sørensen < @.***> ha scritto:

The JMLR paper says this: image.png (view on web) https://github.com/ocbe-uio/BayesMallows/assets/21175639/2c5b40c3-b43c-4faf-9b05-8948393b3b5a

I find it hard to believe we actually used L=40.

— Reply to this email directly, view it on GitHub https://github.com/ocbe-uio/BayesMallows/issues/378#issuecomment-1948188234, or unsubscribe https://github.com/notifications/unsubscribe-auth/AVELG7B5D62UKKPIOCVSP2TYT44ORAVCNFSM6AAAAABDLT6Z3KVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTSNBYGE4DQMRTGQ . You are receiving this because you are subscribed to this thread.Message ID: @.***>

wleoncio commented 6 months ago

Just dropping in to add a direct link to the paper (it's also on the README, FWIW): https://jmlr.org/papers/v18/15-481.html