How to fix the number of iterations for the smoothing phase in SAEM in nlmixr ?

[MorganePhilipp]

Dear @mattfidler,

Is it possible to fix the number of iterations for the smoothing phase in SAEM in nlmixr (equivalent to Monolix's K_2 )?

I found in the guidelines available on r how to fix seed, nBurn, nEm, nmc but I find nothing about how to fix the number of iterations for the smoothing phase.

I would like to fix it also in my code which is currently as follows:
fit <- nlmixr(f, data, est="saem", saemControl(seed=123456, n.burn=5, n.em=300, nmc=5), table = tableControl(cwres=TRUE, npde=TRUE))

Regards, MP

[mattfidler]

Can you provide the link to the monolix documentation? I couldn't find information on Monolix's smoothing phase.

One point, I think yo need to change your saem options to:

fit <- nlmixr(f, data, est="saem", saemControl(seed=123456, nBurn=5, nEm=300, nmc=5), table = tableControl(cwres=TRUE, npde=TRUE))

I can't recall if n.burn works or not.

[MorganePhilipp]

Dear @mattfidler,

Here is the link to the monolix documentation where the informations about the smoothing phase appear: http://monolix.lixoft.com/single-page/?section=population-parameter-estimation-using-saem

Thank you, I did the modifications.

Thank you for your help, Regards, MP

[mattfidler]

Monolix may have changed their K numbering with recent releases, and I don't think they refer to them any longer.

I couldn't find the K_2 in the link you provided. However, I found the discussion of the smoothing phase. I believe nEm is the same as the smoothing phase. The documentation that I was using in monolix was much older than the new online help.

[mattfidler]

I assume this answers your question @MorganePhilipp . Let me know if I'm mistaken.

[MorganePhilipp]

Dear @mattfidler,

Yes, thank you very much for your answer. I'm sorry I thought I already answered you.

Best regards, MP

[mattfidler]

Great! thanks.