LAMARC Changelog
What's new in LAMARC 2.1.10
Sep 5, 2014- FIXED: lam_conv file converter reversed source and destination populations when using custom migration matrices. If you used a custom matrix to set asymmetrical migration rates in any prior version, you will need to re-convert your files and re-run your analysis. All other results are unaffected.
New in LAMARC 2.1.9 (Jan 13, 2014)
- FIXED: multiple severe bugs in inference of population divergence. All analyses involving population divergence should be re-run.
- FIXED: failure of DNA or SNP runs with invariant loci. If your run completed successfully you did not encounter this bug and re-running is not necessary.
- FIXED: lam_conv file converter wrote a multi-population XML file for some single-population cases. If your converted file led to a successful run you did not encounter this bug and re-running is not necessary.
New in LAMARC 2.1.6 (Sep 1, 2011)
- FIXED: major bug with data likelihood calculations for runs modeling recombination. All runs which inferred recombination rate or held it to a non-zero value in versions 2.1.2 through 2.1.5 should be re-run.
- FIXED: calculation of invariant data term for SNP data modeled with a per-base error rate. All 2.1.5 runs which used the data uncertainty model on SNP data should be re-run.
- UPDATED: random number generator to use the Boost library, and added guards against returning values of exactly 0 or 1 for double values. No previous runs need be re-done, but if you had a data set with an unexplained crash, this may fix it.