MemoryError

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Open a large RT Dose file, with a >300 slice CT image set and >30 structures
2.
3.

What is the expected output? What do you see instead?
Expected: standard window, DVH window.
Instead: 

What version of the product are you using? On what operating system?
Version 0.4.1-1, Windows XP

Please provide any additional information below.

Is it possible to increase the memory?  It appeared that it used up to about 
650 MB of RAM and then crashed.  Phil Taddei, phillyla@hotmail.com

System Information and Problem Details:

Platform information:
System: Windows
Version: ('XP', '5.1.2600', 'SP3', 'Multiprocessor Free')
Python Version: 2.6.6
dicompyler Version: 0.4.1-1

dicompyler.dicomparser: DEBUG: Found ROI #4: orig_CTV_lower
dicompyler.dicomparser: DEBUG: Found ROI #6: orig_Brain-GTV
dicompyler.dicomparser: DEBUG: Found ROI #128: 50%IDLpt
dicompyler.dvhcalc: DEBUG: Calculating DVH of 1 brain_stem
dicompyler: ERROR: Unhandled exception: Traceback (most recent call last):
  File "main.py", line 691, in Run
  File "threading.pyo", line 484, in run
  File "main.py", line 438, in LoadPatientDataThread
  File "dicompyler\dvhcalc.pyo", line 21, in get_dvh
  File "dicompyler\dvhcalc.pyo", line 58, in calculate_dvh
  File "dicompyler\dicomparser.pyo", line 560, in GetDoseData
  File "dicom\dataset.pyo", line 398, in _get_pixel_array
  File "dicom\dataset.pyo", line 393, in _getPixelArray
  File "dicom\dataset.pyo", line 357, in _PixelDataNumpy
MemoryError

Original issue reported on code.google.com by phill...@gmail.com on 3 Oct 2012 at 1:31

[GoogleCodeExporter]

On a different machine, Windows 7 with more physical memory, the patient's 
files opened, but when I maximized the window, I got the following fatal error:

Platform information:
System: Windows
Version: ('7', '6.1.7600', '', 'Multiprocessor Free')
Python Version: 2.6.6
dicompyler Version: 0.4.1-1

dicompyler.dvhcalc: DEBUG: Calculating DVH of 125 lungs
dicompyler.dvhcalc: DEBUG: Calculating DVH of 126 breast
dicompyler.dvhcalc: DEBUG: Calculating DVH of 127 skin
dicompyler.dvhcalc: DEBUG: Calculating DVH of 128 50%IDLpt
dicompyler: ERROR: Unhandled exception: Traceback (most recent call last):
  File "P:\python\dicompyler-bdist\dist\baseplugins\2dview.py", line 611, in OnMouseMotion
  File "P:\python\dicompyler-bdist\dist\baseplugins\2dview.py", line 626, in OnLeftIsDown
TypeError: unsupported operand type(s) for -: 'tuple' and 'Point'

Original comment by phill...@gmail.com on 3 Oct 2012 at 4:39

[GoogleCodeExporter]

I think the error above(second post) is relating to a mouse action in the 
2dview.py too soon after changing the drawing state. 

This same tuple/Point error used to occur, with the exe only, when attempting 
to open a patient and double clicking the series in the tree view rather than 
hitting the select button.

I think this might be the case when the 2dview is being redrawn for the 
maximized mode. Could you try to hit ALT+SPACE to bring up the windows menu and 
then press X to select maximize. But you will want to make sure the mouse is 
NOT over the 2dview, so place it down in the start menu to prevent this on 
maximize. That should prevent the tuple error and crash.

Original comment by lstrd...@gmail.com on 4 Oct 2012 at 12:25

[GoogleCodeExporter]

Hi Phil,

lstrdean is correct regarding the mouse action in the 2D view. This was 
resolved in Issue 62.

Regarding the first issue, I have never seen a memory issue before, but granted 
I have not tried a large RT Dose dataset. This may be specific to accessing the 
dose grid via NumPy via pydicom. I will need to investigate what is the root 
cause, as this may be something that may have to be corrected in an upstream 
library.

Thanks for the report.

Original comment by bastula on 4 Oct 2012 at 3:25

[GoogleCodeExporter]

Update:  After I adjusted my mouse movement habits, as instructed, I was able 
to read the RT Dose (and other) files on my Windows 7 workstation.  It has 8 GB 
of memory, but dicompyler only used about 950 MB of memory.  I was unable to 
open the files on my Windows XP workstation, which has less but plenty of 
memory.  Dicompyler only used a few hundred MB of memory before crashing.

Phil

Original comment by phill...@gmail.com on 5 Oct 2012 at 1:12

[GoogleCodeExporter]

Might be able to quickly fix the memory error if the Numpy arrays are split 
into multiple ones, any way to prevent a massive element wise operation. It 
could be a problem with copying an over sized array. This memory error is a 
pain with 32bit Numpy. 

Phil, if you are handy with running Python I believe you can download a newer 
version of Dicompyler which would include the mouse moving fix. However this 
shouldn't affect the  memory limit.

Also, I have worked with oversized RTDose files in the past. However my files 
were created from a simulation system which included too much precision in the 
decimals in printing. After I reduced this, the dose file was reduced my 
hundreds of MB.

Original comment by lstrd...@gmail.com on 5 Oct 2012 at 4:06