Trouble with Masking satellites

[oleg]

Trouble with Masking satellites
> Not enough memory for vector->segment (vector->nsegment elements) !

[Knickohr]

Hi oleg,

how much RAM do you have in your PC ? Do you use Ubuntheli ?

Thomas

[oleg]

Hi Thomas,

configuration of my computer:

CPU: Intel® Pentium® Processor E6700 (2M Cache, 3.20 GHz, 1066 FSB)
RAM: 2GB+2GB+1GB
HDD: 2 TB
...
OS: Ubuntu 11.04 32-bit

[Knickohr]

Hi,

how many files do you process in the pipe ?

Post me the output from

# swapon -s

Thomas

[oleg]

I've set the swap space for your computer about 6.6 GB

[oleg]

Hi,

how many files do you process in the pipe ?

The maximal number of files (science data or "light"), which I processed and stacked into one final mosaic on my PC, was 960 fits (or 8x120).

[oleg]

oleg@oleg:~$ swapon -s
Filename Type Size Used Priority
/dev/sda2 partition 6869136 182108 -1

[Knickohr]

Mosaic ?

Oh dear, 5GB RAM is not enough

Thomas

[oleg]

Hi,

how many files do you process in the pipe ?

I had a problem with Masking satellites when I was processing the 56 fits (7x8)

[Knickohr]

upps

This should work !

OK, we have to wait until Mischa has an idea

Thomas

[mischa]

Hi Oleg,

Trouble with Masking satellites
> Not enough memory for vector->segment (vector->nsegment elements) !

(1) I have seen that before (I think twice)
(2) I have no idea what causes it
(3) Re-create the polygon mask, maybe it goes away. Or park all other images and only try with that one that has the satellite
(4) If nothing helps, do it manually with a different programme: in the SCIENCE/reg directory, do:

fitspolygon -i ../../WEIGHTS/image.weight.fits -p image.reg -o ../../WEIGHTS/tmp.fits -s out
cd ../../WEIGHTS/
mv tmp.fits image.weight.fits

mischa