xscan fails

[cjayb]

mf.detect_bad_chans_xscan(in_fname=in_name[0], set_bad=True)

Traceback (most recent call last):

 File "<ipython-input-105-c0af46da0146>", line 1, in <module>
   mf.detect_bad_chans_xscan(in_fname=in_name[0], set_bad=True)

 File "/usr/local/common/meeg-cfin/stormdb/stormdb/process.py", line 100, in detect_bad_chans_xscan
   raise RuntimeError(errmsg)

RuntimeError: xscan exited with an error, output is:

usage: /neuro/bin/util/x86_64-pc-linux-gnu/xscan-3.0 -f <infile> [options]
'/neuro/bin/util/x86_64-pc-linux-gnu/xscan-3.0 -help' shows available options.

[cjayb]

Still have to figure out how to run xscan: it's resource-heavy so needs to be submitted to queue. But somehow we need to get information on the task completing, and a list of bad chans being available. This is a similar problem to figuring out how far one's maxfilter-runs have got in the queue.

[cjayb]

xscan is not yet production-ready, but you can submit it as a regular string using submit_to_isis (bash) or Maxfilter.submit_to_isis(self, n_jobs=1, fake=False, submit_script=None).

See #10

[sarathykousik]

xscan failed on two control datasets as well but ran well all through the DBS datasets. That was weird to me. Changing the corr didn't really help.

[cjayb]

what does a "fail" look like? anything relevant in the logs?

[sarathykousik]

Log says:
"Program died due to signal 11, all output files were deleted!
Core dump disabled"

[cjayb]

Can you run Maxfilter on the same file without problem, though? With or without autobad? If so, could be an xscan bug, please send path to file so I can share with Elekta

[sarathykousik]

Yes. It runs fine with Maxfilter with autobad OFF and ON. but still fails with xscan.