Converts raw cadence target data from the Kepler space telescope into astronomer-friendly FITS files.
Switch branches/tags
Clone or download
Latest commit 76d48a3 Mar 17, 2018
Type Name Latest commit message Commit time
Failed to load latest commit information.
kadenza bugfix: enable kadenza to work without --collateral set Aug 31, 2017
.gitignore update gitignore Mar 25, 2017
.travis.yml Update .travis.yml Mar 8, 2018
LICENSE Initial commit Sep 27, 2015 Bugfix: add `include_package_data=True` to to resolve #2 Mar 2, 2017 Update Mar 17, 2018
requirements.txt Add tqdm dependency Mar 8, 2017 Add tqdm dependency Mar 8, 2017

Kadenza: Kepler/K2 Cadence Data Reader

Converts raw cadence data from the Kepler spacecraft into astronomer-friendly FITS files.

Travis status MIT license astropy DOI ADS Bibcode

Kadenza enables time-critical data analyses to be carried out using NASA's Kepler Space Telescope by enabling users to convert the "level 0" raw data files into user-friendly Target Pixel Files.

The primary motivation for this tool was to enable the microlensing, supernova, and exoplanet communities to create quicklook lightcurves for transient events and exoplanets which require rapid follow-up.

Kadenza can be used both as a command-line tool or using its Python API. Most users of Kepler/K2 data will not require this tool, because it does not produce fully-calibrated data products.


If you have a working installation of Python on your system, you can download and install the latest version as follows:

$ git clone
$ cd kadenza
$ python install

Kadenza has only been tested under Linux with Python 3 at present.


Example 1: Creating a (sparse) Full Frame Image

For K2 Campaign 9, the data archive at MAST provides access to the raw cadence data files. There is one such file per long cadence, named kplrYYYYDDDHHMMSS_lcs-targ.fits, which provides the pixel counts in that cadence for all those pixels which were pre-selected to be downlinked from the spacecraft (roughly 3% of the 95-megapixel camera). The timestamp that is recorded in the filename refers to the end of the long cadence, which is a period of 0.4904h during which the spacecraft summed 270 exposures of 6.02s each.

To reconstruct two-dimensional images from the cadence data files, we need an additional file called the pixel mapping reference file. This file specifies the (column, row) CCD coordinates for each value in the one-dimensional cadence data arrays. This information is kept in a separate file to avoid having to repeat the pixel coordinates in each cadence file. For K2 Campaign 9a the mapping file you need is called kplr2016068153039-085-085_lcm.fits and can be obtained from the archive.

One you have obtained the raw data, you can then use Kadenza to convert them into two-dimensional images. Once intalled (see below), Kadenza adds the kadenza-ffi tool to your command-line which you simply call as follows:

$ kadenza-ffi cadence-data-file pixel-mapping-file

In our example, we'd execute:

$ kadenza-ffi kplrYYYYDDDHHMMSS_lcs-targ.fits kplr2016068153039-085-085_lcm.fits

This will create a new file called kplrYYYYDDDHHMMSS_kadenza_ffi_raw.fits which is a FITS file with 84 image extensions, each corresponding to the different Kepler CCD channels. The units are counts and unobserved pixels are set to "-1". Timestamps refer to the end of each cadence can be obtained from the filename or the DATE-END keyword in the header.

Example 2: Creating a Target Pixel File

You can also convert all the raw cadence data for a given campaign and target into a Target Pixel File (TPF) file using the kadenza-tpf tool. For example, to create a Target Pixel File for the target with EPIC ID 247520207 observed in Campaign 13, you would first download all the cadence data, e.g. using wget:

$ wget -t1 -e robots=off --no-parent --continue --recursive

Next, create a text file which lists the paths of all the cadence files you downloaded sorted by filename, e.g. using find:

$ find /location/where/you/downloaded/the/raw/cadence/data -name '*lcs-targ.fits' | sort > raw-long-cadence-files.txt

Next, download the long cadence pixel mapping reference file from MAST:

$ wget

And finally, run kadenza-tpf to create an uncalibrated target pixel file:

$ kadenza-tpf --target 247520207 raw-long-cadence-files.txt kplr2017032193929-091-091_lcm.fits

This will create an unofficial Target Pixel File called ktwo247520207-unofficial-tpf.fits.


There are two main caveats to be aware of:

  • The present version does not set all header keywords exactly as they would appear in an official product. In particular, the WCS keywords are untested.
  • This tool does not calibrate the data, it merely serves to transform the raw pixel counts into a FITS format that is similar to, but not at all identical, to the official pipeline products.

Commands provided

$ kadenza-ffi --help
usage: kadenza-ffi [-h] cadence_file pixelmap_file

Turns a raw Kepler Cadence Data file into an uncalibrated Full Frame Image

positional arguments:
  cadence_file   path to the '*_lcs-targ.fits' cadence data file
  pixelmap_file  path to the '*_lcm.fits' pixel mapping reference file

optional arguments:
  -h, --help     show this help message and exit
$ kadenza-tpf --help
usage: kadenza-tpf [-h] [-t [target_id]] cadencefile_list pixelmap_file

Turn raw Kepler Cadence Data into uncalibrated Target Pixel Files (TPF).

positional arguments:
  cadencefile_list      path to a text file that lists the '*_lcs-targ.fits'
                        cadence data files to use
  pixelmap_file         path to the '*_lcm.fits' pixel mapping reference file

optional arguments:
  -h, --help            show this help message and exit
  -t [target_id], --target [target_id]
                        only produce a TPF file for a specific EPIC/KIC


To report bugs and request features, please use the issue tracker or open a pull request.


Kadenza is made available under the MIT License. For details see the LICENSE file.


kadenza was created by Geert Barentsen and José Vinícius de Miranda Cardoso for NASA's Kepler/K2 Guest Observer Office. If this tool aided your research, please consider offering co-authorship to the authors, or at the very least, cite this tool using both the DOI identifier and the ASCL entry using the following BibTeX entry:

  author        = {{Barentsen}, G. and {Cardoso}, J.~V.~d.~M.},
  title         = "{Kadenza: Kepler/K2 Raw Cadence Data Reader}",
  keywords      = {Software },
  howpublished  = {Astrophysics Source Code Library},
  year          = 2018,
  month         = mar,
  archivePrefix = "ascl",
  eprint        = {1803.005},
  adsurl        = {},
  adsnote       = {Provided by the SAO/NASA Astrophysics Data System},
  doi           = {10.5281/zenodo.344973},
  url           = {}