Ticket #10204 (inprogress)
Python rebinning memory leak
Reported by: | Harry Jeffery | Owned by: | Martyn Gigg |
---|---|---|---|
Priority: | major | Milestone: | Backlog |
Component: | Framework | Keywords: | |
Cc: | Blocked By: | ||
Blocking: | Tester: |
Description
From a user:
Unfortunately I have found another issue with Mantid scripting. Using the algorithms natively I can produce the following script history:
Load(Filename='\\isis\inst$\NDXENGINX\Instrument\data\cycle_12_1\ENGINX00185905.raw', OutputWorkspace='ENGINX00185905', LoaderName='LoadRaw', LoaderVersion=3, SpectrumMax=2400) ConvertUnits(InputWorkspace='ENGINX00185905', OutputWorkspace='185905-d', Target='dSpacing', AlignBins=True) Rebin(InputWorkspace='185905-d', OutputWorkspace='185905-bin', Params='0.001')
However, executing the last line in any python shell consumes infinite system memory indefinitely. Executing the Rebin algorithm on its own typically takes 0.1 seconds. The problem persists in the distributions anaconda, canopy and IDLE, and is reproduced in the mantidplot command-line, as well as the python.exe contained in MantidInstall/bin/. This is a new problem on 3.2.0.
I have tested this thoroughly on Ubuntu 14.04 against the latest master and could not reproduce it. I suspect this may be a Windows specific problem. Could someone with a Windows environment try to reproduce this.
Change History
comment:1 Changed 6 years ago by Nick Draper
- Owner changed from Anyone to Martyn Gigg
- Status changed from new to assigned
comment:2 Changed 6 years ago by Martyn Gigg
- Status changed from assigned to inprogress
I can't reproduce this on Windows either. I would be surprised if this is truly a bug in Rebin. Almost every time someone reports a problem with Rebin, it ends up being something else entirely.
I'll get in touch with the user directly to try and help figure this out.