findThreshold seems to be stuck forever

[Julian Zhou]

For only ~1,100 or so sequences, findThreshold() seems to be stuck at Iteration 5 - Fitting gamma-gamma for good.

Example file attached.

library(shazam)

load("distNthresh_HD07_Naive.RData")

threshold.gmm <- findThreshold(data=cur.db$DIST_NEAREST, method="gmm",
                                       model="gamma-gamma", cutoff="optimal", 
                                       sen=NULL, spc=NULL, progress=T)

[Julian Zhou]

Talked to Nima (quite a while back actually; forgot to update this issue). This could happen by design (i.e. how the algorithm works/is limited) if the data really isn't bi-modal.

[Julian Zhou]

Pro-tip from Nima: this is why one should still always perform manual inspection of the distToNearest histograms before calling findThreshold().

[Julian Zhou]

^ P.s. He also agreed to add that tip to the vignette.