migrate 4.4.4 problem

[Peter Beerli]

Dear all,
we found a problem with the distributed version migrate 4.4.4 (but not with others) that caused an underestimate of mutation-scaled immigration rates.
The error is most prominent if the immigration rates are huge (e.g. the mutation-scaled population sizes are very small) and most of the immigration events are happening in the far past (near the most recent common ancestor). It also seems that as a result of this problem, the PDF plots of the overall loci immigration rates returned blank histograms.

I have removed the 4.4.4 version from the repository and add a replacement with the fix as 4.4.5. I suggest to use to use version 5.x in general;
and, …
I am convinced that all software that is sufficiently large contains problems, the hope is simply that these problems are minor.

If this error caused problems, I apologize,
Happy new year,
Peter


Peter Beerli
Scientific Computing, FSU, Tallahassee
bee...@fsu.edu, beerli...@gmail.com