| title | tags | authors | affiliations | date | bibliography | |||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
dfitspy: a dfits|fitsort implementation in python |
|
|
|
28 January 2019 |
paper.bib |
The FITS format (Flexible Image Transport System) [@Hanisch:2001] is a widely used format to store astronomical data. It is used to store a lot of different types of data such as 1D or 2D spectra, 3D data cubes. It has been developed in the late 1970 to reach its final form almost two decades ago. FITS files are built with two components. The data themselves are stored as tables and contains any types of data. A header is built containing set of keywords-value pairs aiming at describing the data themselves.
Accessing and displaying metadata inside FITS files is important in order to get an overview of their content properties without having to read the data themselves. It is particularly useful when dealing with large amount of files at once. Tools have been already publicly available for years with the dfits and fitsort algorithms (the documentation is available here https://www.eso.org/sci/software/eclipse/eug/eug/node8.html). The main limitation is that they are stand-alone programs useable only in a terminal. They can not be used natively inside another program.
The python module presented in this paper, dfitspy, is a project that migrates the main dfits
and fitsort capabilities to python. It is a metadata searcher/displayer for FITS files.
As dfits and fitsort, dfitspy is able to display in the terminal the result of a metadata
search and is able to grep certain values of keywords inside large samples of files.
Therefore it can be used directly with the command line interface. Nevertheless,
dfitspy can be, and it is its strength, imported as a python module and the user can
use these functionnalities inside another python code or the python interpretor.
A command line interface has been included in dfitspy so it can be used as a Terminal command. A typical command is:\
dfitspy -f Test_data/* -k author,number,type --grep 2dspec\
This command will search in all the FITS file present in the Test_data directory. dfitspy will search for three keywords in the header: author, time and type. Finally, the terminal will display only the file where 2dspec is in the requested keyword values. The terminal output is similar to the dfits|fitsort combination. It displays, in a column fashion, each file with the requested keyword its corresponding values:
\newpage
filename author number type
---------- ------------ -------- ------
file1.fits R. Thomas 49098.26 2dspec
file2.fits R. Thomas 79098.26 2dspec
file3.fits R. Thomas 69198.26 2dspec
file4.fits R. Thomas 79498.26 2dspec
file5.fits R. Thomas 89098.26 2dspec
file6.fits R. Thomas 79498.26 2dspec
To be used as a Python module, dfitspy must be imported. Then a set of command have to be used in order to produce the final list of filenames/keywords/values. In short, three main commands must be used:
First of all, import the module:
import dfitspy
Then, the files must be gathered:
listfiles = dfitspy.get_files(['all'],'Test_data/')
And the list of keywords must be prepared, and eventually the grepping values:
listkeys = ['author', 'number', 'type']
grepping = ['2dspec']
Finally, we can fitsort the files and eventually grep.
fitsortgrep = dfitspy.dfitsort(listfiles, listkeys, grepping)
The final output is stored as a dictionnary of files for which each keywords/values is given. It can also be displayed in the same way as for the terminal output (see above).
dfitspy is a GPL licensed software and the source code is available at https://github.com/astrom-tom/dfitspy. The full documentation is available at https://astrom-tom.github.io/dfitspy/build/html/index.html .
The author would like to thank the Journal of Open Source Software to give the opportunity to researchers to publish their softwares and to the referee of this paper for helpful comments.