python 2/3 compatible
Requires numpy, scipy for the main module "nonlinear_mass_faster.py," plus "taylor_coeffs.bin" and "fiducial_rnl.txt"
Requires camb to generate the power spectrum files
The main module is "nonlinear_mass_faster.py" and an example is shown in "example.ipynb." One can simply use "nonlinear_mass_faster.py" + "taylor_coeffs.bin" + "fiducial_rnl.txt" to either fit the polynomial coefficients, or use the Taylor series in the default Planck18 cosmology.
We also include various helper scripts in case the user wants to re-generate the Taylor series about a cosmology of their choosing.
"make_pklin.py" and "make_pklin_all_cosmos.py" generate the default linear power spectra, and the linear power spectra for tests and the Taylor series.
"compute_taylor_coeffs.py" creates the "taylor_coeffs.bin" file from the linear power spectra.
"tabulate_fiducial_rnl.py" tabulates RNL in the fiducial cosmology (Planck18), needed for setting the fitting range.
Finally, we include two scripts that replicate the key tests in Krolewski & Slepian 2019: "accuracy.py" and "timing_test.py" run the accuracy and timing tests.