MBProj2D

2D forward-fitting projection code for fitting images of clusters of galaxies. This work builds upon the ideas in MBProj2 (see https://github.com/jeremysanders/mbproj2 and https://ui.adsabs.harvard.edu/abs/2018MNRAS.474.1065S). It differs from MBProj2 in that it fits images, rather than profiles, and allows multiple cluster and background components to be simultaneously fitted.

This code is in development, so please be aware that the interface can be unstable. More documentation can be found at https://mbproj2d.readthedocs.io/en/latest/

Copyright Jeremy Sanders (2020)

License: LGPLv3

Requirements:

• Xspec (https://heasarc.gsfc.nasa.gov/docs/xanadu/xspec/)
• python3
• numpy
• scipy
• cython
• pyfftw
• h5py
• emcee
• astropy
• nlopt (optional, for better fitting)

Installation instructions using pip:

python3 -m pip install astropy cython emcee h5py numpy pyfftw scipy wheel
python3 -m pip install git+https://github.com/jeremysanders/mbproj2d

If using conda, the following package is also required for Linux:

conda install -c conda-forge gxx_linux-64

Usage notes:

• If using a PSF model, I suggest fitting a larger region of the sky than necessary, but masking out the edges. Do not zero the exposure map in these regions. The PSF modelling uses a FFT convolution, so the model will wrap around at the edges. Please note that the pad=N option in Image() or imageLoad will automatically zero-pad images with N masked-out pixels.

• Input images must have even numbers of pixels on each axis if optimal_size not set