6. Issues and Limitations
6.1. Memory Usage
Some primitives use a lot of RAM memory and they can cause a crash. Memory
management in Python is notoriously difficult. The DRAGONS’s team is constantly
trying to improve memory management within astrodata
and the DRAGONS recipes
and primitives. If an “Out of memory” crash happens to you, if possible for your
observation sequence, try to run the pipeline on fewer images at the time,
like for each dither pattern sequence separately.
Then to align and stack the pieces, run the alignAndStack
recipe:
$ reduce @list_of_stacks -r alignAndStack
6.2. Double messaging issue
If you run Reduce
without setting up a logger, you will notice that the
output messages appear twice. To prevent this behaviour set up a logger.
This will send one of the output stream to a file, keeping the other on the
screen. We recommend using the DRAGONS logger located in the
gempy.utils.logutils
module and its
config()
function:
1from gempy.utils import logutils
2logutils.config(file_name='gmos_data_reduction.log')
6.3. Astropy warnings
You might see some warning messages from AstroPy that are related to the header of the images. It is safe to ignore them.