reduceccd

Basic Image CCD reduction using CCDPROC package

reduceccd is a small pipeline that performs basic automatic image CCD reduction on a set of files:

• Create master bias
• Create master flat for each filter
• Performs bias and flat fielding correction of the science images
• Cosmic ray removal
• Sky subtraction
• Align and combination of images of the same filter

Installation instructions

The following scripts require these packages to be installed in order to run:

• numpy
• scipy
• astropy
• ccdproc
• photutils
• skimage

To install the package type in a terminal:

pip install https://github.com/rgbIAA/reduceccd/archive/master.zip

Or download the package and inside the directory type:

python setup.py install

Instructions

If the images contain the basic information in the header, a standard reduction script would be something like:

import reduceccd as rc
import warnings
import logging

logging.disable(logging.INFO)
warnings.simplefilter('ignore')

# Your raw data
path = '/User/me/route/to/my/raw/data/'
# Output directory
dout = 'redudir/'

# If you want to reduce all filters, leave "None"
# If you just want to reduce a few filters, provide a list or a string
filters = None # ['Ha6563', 'Ha6607', 'V', 'R', 'B']
# Trim section (None otherwise)
fits_section ='[:,:995]'
# Calculate uncertainty and propagate error (requires BOTH gain and readnoise)
error = True 

# CCD info
gain = 1.5 # +- 0.02 e / ADU
readnoise = 8.23 # +- 0.12 e rms

# If you want to reduce all objects, just write "None" 
# If you just want to reduce a particular set of objects, provide a list or string (one object)
objects = None # 'ngc7465'

rc.reduceNight(path, filters, fits_section=fits_section, dout=dout, gain=gain, readnoise=readnoise, 
	create_flat=True, create_bias=True, sky=True, correct_images=True, objects=objects, 
	align=True, error=error)

By default, reduceNight looks into the header of the images to sort the images by type in the following way:

• Bias: keyword imagetyp = 'BIAS'
• Flat: keyword imagetyp = 'FLAT'
• Object: keyword imagetyp = 'LIGHT'

You can modify the way reduceNight looks for the different types of images. Let's say one night you have the keyword imagetyp = 'FLAT' in the header of some flats, and in others you have the string "flat" in the name of the file, but NOT in the header. You can give an OR search command in the form of a list of dictionaries:

dfilter_flat = [{'imagetyp':'FLAT'}, {'find':'flat'}]
rc.reduceNight(path, filters, fits_section=fits_section, dout=dout, gain=gain, dfilter_flat=dfilter_flat)

reduceNight will take as flats files that either have the keyword imagetyp = 'BIAS' in the header OR files that contain the string "flat" in their names ('find' is the dictionary keyword to search for a string in a file name).

You can also provide AND instances for searching files just adding more keywords in the SAME dictionary:

dfilter_flat = {'imagetyp':'FLAT', 'find':'flat'}
rc.reduceNight(path, filters, fits_section=fits_section, dout=dout, gain=gain, dfilter_flat=dfilter_flat)

In this case, reduceNight will consider flats files that have the keyword imagetyp = 'BIAS' in the header AND files that contain the string "flat" in their names.

There is also a dfilter_bias and dfilter_images for bias and science images, respectively.

Sky subtraction

There are 3 sky subtraction methods implemented in reduceccd: percentile, mask, and 2d.

• percentile computes the median value of the image within the percentile values introduced by arguments lower (default: 1.) and upper (default=95.)
• mask relies on photutils. Read their documentation to see the available parameters
• 2d relies on photutils. Their background classes can be chosen by the bkg_method argument:
• mean: MeanBackground
• median: MedianBackground
• mode: ModeEstimatorBackground
• mm: MMMBackground
• sextractor: SExtractorBackground
• biweight: BiweightLocationBackground

All these options can be set in reduceNight with the argument dict_sky:

dict_sky = {'method':'2d', 'bkg_method': 'mode'}
rc.reduceNight(path, filters, dict_sky=dict_sky)

Alignment

The alignment of the images is done by means of skimage package. Sometimes, the alignment method can go bananas when the field is too large. For alignment purposes, the field can be reduced with the zoom argument:

dict_align_combine = dict(zoom=2, center=(230, 300))
rc.reduceNight(path, filters, dict_align_combine=dict_align_combine)

where the image will be cropped by half from the center of the image by default. The center can be controlled by the center argument.