search

Irrationone / cellassign

Automated, probabilistic assignment of cell types in scRNA-seq data
Other
191 stars 79 forks source link

upper bound for qdiff in the event pass optimum #30

Closed nceglia closed 5 years ago

nceglia commented 5 years ago

I ran into some situations in the pipeline where the set parameters would go past the minimum and start climbing, so this would prevent it from climbing too far.

Irrationone commented 5 years ago

Should be fixed more directly by #31. Issue was with underflow in theta_log_prob that would occur somewhat randomly in cases where particular cell types don't exist.

If that doesn't work, let me know and I'll revisit this.