VPC Error for a Bayesian fit object

Diagnostics
1

Hi there. I am unable to generate a PPC from a Bayesian fit object using the default vpc() and vpc_plot() functions. The error message indicates that these functions do not support fit objects with censored_latent(). Could you kindly suggest a workaround? Many thanks.

2

Hi Ayorinde - are you on Pumas 2.6?

Best,

Vijay

3

If you can post the error message it would be very helpful. Thanks

4

We looked at it internally, and you are right. The VPC workflow that first simulates and then calculates the VPC quantities does not currently support censored observations (also does not yet support prediction correction). This is scheduled for Pumas v2.8.

Depending on what you need to do exactly, it should be possible to work with the DataFrames constructed from the simulations and get what you need. Let us know if you need help to accomplish this.

5

Many thanks Patrick - I’ll do stuffs manually then. @Vijay Yes, I’m on version 2.6.

1 Like