Chroma is written entirely in Python. Once downloaded, there is nothing to install; all python scripts are intended to run from the directory in which they are unpacked.
Chroma requires several python libraries to run, although not all libraries are required for all analyses. The basic set of dependencies, required everywhere, are
which are sufficient to enable one to evaluate chromatic biases and make plots.
With the added dependency
one can also use machine learning algorithms to predict the size of chromatic biases given photometric data, which can then lead to a correction mitigating the bias. This step requires a catalog on which to run, which can either be built using the LSST catalogs framework with the script supplied, or alternatively can simply be downloaded here.
Finally, to test analytic results in simulated images, the following packages are required:
Actually, astropy is a nice tool to have around for the earlier analyses too, since, when available, we use its console module to create a nice looking progress bar for steps that take a while.
The primary directory structure of chroma is the following:
(There are additional directories lurking, but the above are the ones that are most interesting and up-to-date).
The data/ directory contains files representing a small number of SEDs, and the LSST filter bandpasses.
The chroma/ directory contains a few useful library modules:
Scripts that produce output reside in the bin/ directory. We will go through the various bin/ scripts under the results tab above.