tree.shared is hanging when I include the subsample option. Runs fine with no subsample specified.
What version of mothur are you running?
What do you mean by hanging? How many samples are you trying to analyze? What is your subsample size?
1.30.1
tree.shared stops responding. Subsample size is 750.
Extract from logfile:
mothur > tree.shared(calc=thetayc-jclass, subsample=750)
Using final.an.shared as input file for the shared parameter.
Using 8 processors.
10S contains 382. Eliminating.
11S contains 631. Eliminating.
28S contains 560. Eliminating.
2S contains 356. Eliminating.
37S contains 407. Eliminating.
40S contains 543. Eliminating.
47S contains 523. Eliminating.
5S contains 530. Eliminating.
0.015
[no response from this point so used Ctrl-C to abort]
quitting command…
mothur > tree.shared(calc=thetayc-jclass)
Using final.an.shared as input file for the shared parameter.
Using 8 processors.
0.015
Output File Names:
final.an.thetayc.0.015.tre
final.an.jclass.0.015.tre
[ran fine without subsample]
How many groups do you have in your shared file?
99
How long did you wait before killing it?
About 2 hours.
How long does the command take without subsampling? With subsampling the command should take more than 1000 times longer, because the calculations are preformed 1000 times, plus the time to subsample and find the averages.
It takes seconds without subsampling. I hadn’t realised there would be a 100-fold difference. I’ll leave it to run overnight.
Thanks
William
It did indeed complete eventually. I had not realised that there would be such a difference in execution time when the subsample option is included. Thank you for your help.