README.md

CHIMExIllustrisTNG

Intended to store code and mock catalogs and intensity maps related to Polzin et al. (2024).

There are 3 (pre-beam convolution, beam-convolved, and beam-convolved + padded) intensity maps at each of z = 1 and 2 in the maps directory of this repository for each of the following catalogs:

1. all galaxies (field galaxies and cluster galaxies)
2. mass-selected galaxy clusters
3. radius selected galaxy clusters

These maps, and the beams, are saved as .npy files, which can be read in the following way:

import numpy as np
filename = np.load('filename.npy')

Additionally, there are pre-permutation catalogs corresponding to each of these intensity maps in the catalogs directory and code used in the creation and analysis of mock intensity maps and in the permutation of the volume/catalogs in the analysis directory. Ideally, this means that the code + catalogs are enough to do similar analyses without downloading tens of GBs of TNG300 group catalogs (though you will still need to download the Molecular and atomic hydrogen (HI+H2) galaxy contents catalog at z = 1 and 2, which are a total of ~4 GB).

We offer a guide to what changes should be made to the code for different use cases.

How to modify code + catalogs to match your work:

• Since everything should be accessible in the target catalogs we provide here + the Molecular and atomic hydrogen (HI+H2) galaxy contents catalog, the only reason you should need to use the TNG300 group catalogs in a repeated analysis are if you are changing the cluster selection criteria. If you are changing the HI-H₂ model, you can simply overwrite the pre-permutation catalogs with new M_HI values from the "Molecular and atomic hydrogen galaxy contents catalog" and re-run the tiling/stacking. If you are changing the cluster selection criteria, then you will need to modify check_M200 and the condition on R200 in chime_mock_catalog.py and re-run build_HI_cats.
• If you are using this code/these catalogs with another experiment, you will need to feed chime_mock_analysis.py new arrays representing your beam, will want to modify the pixel resolution in make_bins, tile, and make_stack_map, swapping out 5.3 and 5.7 (arcmins) for the relevant values in your case, and will want to change the global "freq_width" to correspond to your instrument's spectral resolution. Similarly, the stacks being of shape (69, 65) is due to the angular resolution of the pixels, so you will also want to change amount of padding in make_stack_map, and the pad conditions in stack and nsamp_stack. (You should be able to find+replace all for the following values 5.3, 5.7 -- for the pixel resolution, 34, 32 -- for the padding on the stacks, and 69, 65 -- for the stack dimensions to easily align with another instrument's specifications.) At minimum, you will also want to recompute the beam FWHM and f_nt (code at the end of chime_mock_analysis.py) to replace "sigma" in single_gauss_0 and single_gauss_1 and the global variables "fnt" and "fff" (the latter of which is also going to be catalog dependent), and you will likely need to recalibrate the mass recovery to suit your instrument and its beam.
• In any case, chime_make_figures.py should be used as a guide as and should not be run as a separate script as it relies on variables defined in chime_mock_analysis.py. (The figures generated are from the original submitted version of the paper -- the original Figure 1 does not appear in the later published version, and the histograms in the original Figures 7 and 8 have been rotated 90 degrees in the final version. All of the salient code is still in this script, but if you need the updated plotting code for any reason, please get in touch, and I will send it to you.)