MixedModels residuals and post-hocs

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
4 messages Options
Reply | Threaded
Open this post in threaded view
|

MixedModels residuals and post-hocs

Carlos Familia
Hello,

I was wondering if it is possible to take the residuals from a lmm model and if there is any lsmeans package for post-hocs in julia?

Best regards,
Carlos Família

--
You received this message because you are subscribed to the Google Groups "julia-stats" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
For more options, visit https://groups.google.com/d/optout.
Reply | Threaded
Open this post in threaded view
|

Re: MixedModels residuals and post-hocs

Douglas Bates
On Friday, June 17, 2016 at 4:26:35 AM UTC-5, Carlos Familia wrote:
Hello,

I was wondering if it is possible to take the residuals from a lmm model and if there is any lsmeans package for post-hocs in julia?

There is no lsmeans package that I know of and it is highly unlikely that I will ever write one, but someone else may.

As for residuals, like many other properties of mixed-effects models, you need to be careful how you define them.  You need to decide if the random effects are included in the fitted values or not. Perhaps it is better to define a fitted method for LinearMixedModel, which can then be the basis of a residuals method.

Could you open an issue in the MixedModels repository?  It is probably better to have this discussion there.

The reason that I say it is unlikely I will write any lsmeans code is because lsmeans are a bizarre SAS construction from the days when we stored data on magnetic tape and programs on punched card decks.  It would take so long to get a single model fit that users wanted every possible inference or diagnostic value calculated from the model, whether or not they made sense.

--
You received this message because you are subscribed to the Google Groups "julia-stats" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
For more options, visit https://groups.google.com/d/optout.
Reply | Threaded
Open this post in threaded view
|

Re: MixedModels residuals and post-hocs

Carlos Familia
Thank you for your answer, and sorry for the late reply.
I will open an issue in the MixedModels repository.


On Friday, June 17, 2016 at 9:52:05 PM UTC+1, Douglas Bates wrote:
On Friday, June 17, 2016 at 4:26:35 AM UTC-5, Carlos Familia wrote:
Hello,

I was wondering if it is possible to take the residuals from a lmm model and if there is any lsmeans package for post-hocs in julia?

There is no lsmeans package that I know of and it is highly unlikely that I will ever write one, but someone else may.

As for residuals, like many other properties of mixed-effects models, you need to be careful how you define them.  You need to decide if the random effects are included in the fitted values or not. Perhaps it is better to define a fitted method for LinearMixedModel, which can then be the basis of a residuals method.

Could you open an issue in the MixedModels repository?  It is probably better to have this discussion there.

The reason that I say it is unlikely I will write any lsmeans code is because lsmeans are a bizarre SAS construction from the days when we stored data on magnetic tape and programs on punched card decks.  It would take so long to get a single model fit that users wanted every possible inference or diagnostic value calculated from the model, whether or not they made sense.

--
You received this message because you are subscribed to the Google Groups "julia-stats" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
For more options, visit https://groups.google.com/d/optout.
Reply | Threaded
Open this post in threaded view
|

Re: MixedModels residuals and post-hocs

Carlos Familia
In reply to this post by Douglas Bates
So what options are there for evaluating the significance of the main effects, and for diagnostics within the julia's MixedModels package?

On Friday, June 17, 2016 at 9:52:05 PM UTC+1, Douglas Bates wrote:
On Friday, June 17, 2016 at 4:26:35 AM UTC-5, Carlos Familia wrote:
Hello,

I was wondering if it is possible to take the residuals from a lmm model and if there is any lsmeans package for post-hocs in julia?

There is no lsmeans package that I know of and it is highly unlikely that I will ever write one, but someone else may.

As for residuals, like many other properties of mixed-effects models, you need to be careful how you define them.  You need to decide if the random effects are included in the fitted values or not. Perhaps it is better to define a fitted method for LinearMixedModel, which can then be the basis of a residuals method.

Could you open an issue in the MixedModels repository?  It is probably better to have this discussion there.

The reason that I say it is unlikely I will write any lsmeans code is because lsmeans are a bizarre SAS construction from the days when we stored data on magnetic tape and programs on punched card decks.  It would take so long to get a single model fit that users wanted every possible inference or diagnostic value calculated from the model, whether or not they made sense.

--
You received this message because you are subscribed to the Google Groups "julia-stats" group.
To unsubscribe from this group and stop receiving emails from it, send an email to [hidden email].
For more options, visit https://groups.google.com/d/optout.