Runing slowly and reporting errors using the Epi tree prior in BEAST v 2.7.8

[Jie Zha]

Hi,

I am trying to use the Epi tree Prior in BEAST v 2.7.8 for epidemic analysis, which uses the Particle Filter method for finding a good starting point relatively efficiently. However, although the my analysis runing was not aborted due to some error issue, but running feedback the error message as below:

Warning: removing transform class beast.base.inference.operator.kernel.Transform$LogConstrainedSumTransform because it should have at least 2 dimensions but it has 0

And here is the end of the stdout file:

Sample posterior likelihood prior 0 -26357.8657 -25327.2572 -1030.6084 --

Particle weight is NaN - should not happen!
Particle weight is NaN - should not happen!
Particle weight is NaN - should not happen!
…

And the other issue is the runing was very slow.

I attached my xml file below, for you check the source of the issues.

https://github.com/user-attachments/files/22551650/GII17P17-EPI-SIR.xml

Best,

Jie

[Timothy Vaughan]

Hi Jie,

I see you've also opened an issue on the EpiInf repository.  Seeing as this is related exclusively to that package, let's continue this discussion there.

Regards,

Tim