flowClust with prior: "randomStart>0 has no effect when using a prior"
1
0
Entering edit mode
pavitrarc ▴ 20
@pavitrarc-7234
Last seen 2.7 years ago
United States/Seattle/Fred Hutch

I'm using flowClust to identify cell subsets in flow data. I use the untreated samples to obtain a prior, which works well. However, when running flowClust on the remaining samples with the prior, I get this message: 

"randomStart>0 has no effect when using a prior. Labels are initialized from the prior.
Using the parallel (multicore) version of flowClust with 4 cores"

I've tried setting different values for randomStart but it didn't help.

It's not a problem per se.. everything still works fine and the clustering looks good but it just clutters up the console (ocd, I know!). 

Thoughts?

flowclust • 1.3k views
ADD COMMENT
1
Entering edit mode
Greg Finak ▴ 150
@greg-finak-4594
Last seen 5.6 years ago
(Private Address)

Set randomStart=FALSE to eliminate this warning.

 

ADD COMMENT
0
Entering edit mode

Hmm, I tried that but I still get the warning. This was my flowClust call:

prior<-flowClust2Prior(flowClust(all_data[[which(phenoData(all_data)$promoter=='none')]],
                                 varNames=c('FSC-A','SSC-A'),
                                 K=numclusters,
                                 nu=Inf,
                                 level=outlierlev,
                                 min=minvals,
                                 max=maxvals),
                       kappa=1);

res_all<-fsApply(all_data,function(x)flowClust(x,varNames=c('FSC-A','SSC-A'),
                                               K=numclusters,
                                               level=outlierlev,
                                               min=minvals,
                                               max=maxvals,
                                               randomStart=FALSE,
                                               prior=prior,
                                               usePrior='yes'));

ADD REPLY
1
Entering edit mode

Thanks, 

Turns out there was a bug in the code throwing the warning unnecessarily. I've pushed a fix, and it should appear some time tomorrow. Now setting randomStart=FALSE or 0 should suppress the warning properly.

Greg

ADD REPLY
0
Entering edit mode
Works great. Thank you!
ADD REPLY

Login before adding your answer.

Traffic: 718 users visited in the last hour
Help About
FAQ
Access RSS
API
Stats

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.

Powered by the version 2.3.6