search

drizopoulos / GLMMadaptive

GLMMs with adaptive Gaussian quadrature
https://drizopoulos.github.io/GLMMadaptive/
60 stars 14 forks source link

Possibility to evaluate marginal log-likelihood #35

Closed sebastianueckert closed 3 years ago

sebastianueckert commented 3 years ago

Hi Dimitris,

Thanks a lot for your work on this very useful package.

I would like to use GLMMadaptive to calculate a cross-validated log-likelihood, i.e., to estimate model parameters using a subset of the observations and then evaluate the marginal log-likelihood for the hold-out set. I was, therefore, wondering if it is possible to do the integration in GLMMadaptive without any optimization?

Sebastian

drizopoulos commented 3 years ago

Hi Sebastian,

Yes, this is possible by setting iter_EM and iter_qN_outer to zero.

Dimitris

—— Professor of Biostatistics Erasmus Medical Center Rotterdam The Netherlands

Από: Sebastian Ueckert @.> Στάλθηκε: Tuesday, March 30, 2021 12:27:56 PM Προς: drizopoulos/GLMMadaptive @.> Κοιν.: Subscribed @.***> Θέμα: [drizopoulos/GLMMadaptive] Possibility to evaluate marginal log-likelihood (#35)

Hi Dimitris,

Thanks a lot for your work on this very useful package.

I would like to use GLMMadaptive to calculate a cross-validated log-likelihood, i.e., to estimate model parameters using a subset of the observations and then evaluate the marginal log-likelihood for the hold-out set. I was, therefore, wondering if it is possible to do the integration in GLMMadaptive without any optimization?

Sebastian

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHubhttps://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fdrizopoulos%2FGLMMadaptive%2Fissues%2F35&data=04%7C01%7Cd.rizopoulos%40erasmusmc.nl%7C38756056c37640479d0508d8f3667cb2%7C526638ba6af34b0fa532a1a511f4ac80%7C0%7C0%7C637526968788836760%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=lQcxuR4MjqFKRc%2Bk%2F4z5wqG%2FQd2R4y4%2FOK7zjBHZ%2Beg%3D&reserved=0, or unsubscribehttps://eur01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgithub.com%2Fnotifications%2Funsubscribe-auth%2FADE7TT6O24DBNSQZMLFI23DTGGRSZANCNFSM42BXE64A&data=04%7C01%7Cd.rizopoulos%40erasmusmc.nl%7C38756056c37640479d0508d8f3667cb2%7C526638ba6af34b0fa532a1a511f4ac80%7C0%7C0%7C637526968788846712%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C1000&sdata=dJqSWZ6RXaQ1qm0SWlzseNbPXquLGkjIslK%2FvBLK8E4%3D&reserved=0.

sebastianueckert commented 3 years ago

Thanks a lot!