dadi-cli --global-optimization outputs theta = 0.0

[Zoe Meziere]

Hi Ryan,

I have been playing with dadi-cli and very much enjoying the ease of a command-line implementation of dadi!!

My issue is with the output of InferDM when using --global-optimization. As you can see bellow, theta is always 0.0. I don't have this issue when not using --global-optimization.

Is that a problem on my end, and if so, how can I fix it?

Thanks a lot!

Zoe

# Top 100 results

# Log(likelihood) nu1 nu2 T m12 m21 misid theta

-452.7427358272914 0.24583634013575154 1.7646849432488456 0.060019616981904854 0.6501499117470967 0.5743160120726214 0.0

-452.87231032065546 0.26294231502672144 1.8179395160633591 0.06313426623381065 0.30014335620213645 0.8923005673632931 0.0

-452.930250722441 0.2632365369166945 2.1842799429652207 0.06321203230751461 0.3447013522986432 0.6099166432093489 0.0

-453.46579426497476 0.23117105361242454 1.9689548016113496 0.055509250061449605 0.7521878523975937 0.006097868880653796 0.0

 

[Ryan Gutenkunst]

Hello Zoe,

I’m glad you’re enjoying using dadi-cli. You’ve almost certainly found a bug. We’ll track it down and update you.

Thanks!
Ryan

> --
> You received this message because you are subscribed to the Google Groups "dadi-user" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to dadi-user+...@googlegroups.com.
> To view this discussion on the web visit https://groups.google.com/d/msgid/dadi-user/1c457a0a-d86b-4d47-a9a4-b7327fb453dbn%40googlegroups.com.

[Travis Struck]

Hi Zoe,

dadi-cli has been updated (version 0.9.3), so the proper theta0 will be recorded in the output file.

Travis

[Zoe Meziere]

Great, thanks a lot Travis!