covariates with different extrapolation method

[motyocska]

Hi Matt,

thought I would ask: is it possible to specify covariates with different extrapolation method? Ie have weight for example as locf and creatinine as linear within the same model? It sems that this is possible if I do the extrapolation outside of the event table then add et$cov to table that is already extrapolated with the specific method but what I would like is to have rxSolve do the extrapolation work instead...

thanks,

Andras

[mattfidler]

It currently does not work though I thought about adding a few times.

You can use evid=2 to kind of do this but it still is not quite the same.

[billdenney]

My method for linear interpolation is to do it manually by adding 3 columns to the dataset (4 columns total):

• Current covariate value
• Current time (already there as your "time" column)
• Next covariate value
• Next covariate time

Then, I manually interpolate.

My guess is that you may already do this, but hopefully it helps someone else.

[motyocska]

thank you Guys!

[mattfidler]

Since I have been changing covariate handling with iCov, this is probably a good time to change this;

From the release notes:

• Now you can specify the interpolation method per covariate in the model:

  • linear(var1, var2) says both var1 and var2 would use linear interpolation when they are a time-varying covariate. You could also use linear(var1)

  • locf() declares variables using last observation carried forward

  • nocb() declares variables using next observation carried backward

  • midpoint() declares variables using midpoint interpolation

Once #749 is merged you could use the develpment branch to specify different interpolation methods with whatever model you want.

[motyocska]

thanks Matt! Loking forward to try it! Andras Farkas, Pharm.D  On Saturday, August 3, 2024 at 04:05:43 PM EDT, Matthew Fidler @.***> wrote:

Closed #730 as completed via #749.

— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>

[mattfidler]

Great @motyocska