CP2.1.0 command line question

Hey team,

Congrats on the 2.1.0 upgrade, great work. One question: I’m running CP headless and a pipeline that works for me without problems in 2.0 fails in 2.1.0 with the following error:

Version: 2014-01-24T15:16:45 0c7fb94 / 20140124151645
Uncaught exception in CellProfiler.py
Traceback (most recent call last):
  File "CellProfiler.py", line 226, in main
  File "CellProfiler.py", line 717, in run_pipeline_headless
  File "cellprofiler\pipeline.pyc", line 1284, in save_measurements
  File "cellprofiler\pipeline.pyc", line 291, in add_all_measurements
  File "numpy\core\fromnumeric.pyc", line 1800, in amax
ValueError: zero-size array to ufunc.reduce without identity
stopping worker thread  0
stopping worker thread  1
stopping worker thread  2
stopping worker thread  3
stopping worker thread  4
stopping worker thread  5
stopping worker thread  6
stopping worker thread  7
stopping worker thread  8
stopping worker thread  9
stopping worker thread  10
stopping worker thread  11
stopping worker thread  12
stopping worker thread  13
stopping worker thread  14
stopping worker thread  15
stopping worker thread  16
stopping worker thread  17
stopping worker thread  18
stopping worker thread  19
stopping worker thread  20
stopping worker thread  21
stopping worker thread  22
stopping worker thread  23
stopping worker thread  24
Exiting the JVM monitor thread

where my command (which works for 2.0) looks something like:

Any clue? I have imported my 2.0 pipeline into 2.1.0, tested it in GUI mode, and exported it as “pipeline.cppipe” which I am trying to run in headless mode here. Should I have modified the pipeline further?

Also, the 24 worker threads won’t be used for heavy computation right? CP2.1.0 in headless mode should be optimized for single-core use?

Thanks for your help!

Best,
Berend

Hi Berend,

Did it fail on the first image cycle? (It looks like it did from the stack trace, but I want to make sure…). Regardless, would you be able to post the pipeline (both 2.0 and 2.1) and the associated image set?

Thanks,
-Mark

Hi Berend,

Re: multiple core use in headless mode, you are correct (with one caveat). Please see this post from Lee: Multithreaded CellProfiler on a cluster

And this fix may have fixed the headless submission error. Would you test using a trunk build? github.com/CellProfiler/CellPro … TrunkBuild

Thanks,
David