Welcome to GeoFPTax’s documentation!
While this package is focused on the JAX implementation of the Geo-FPT code, it also includes
the original C implementation for testing purposes. The caveat being that the
JAX implementation resorts to 2 1D trapezoidal rule integrations instead of the quadrature used in the C code, this makes the result
slightly dependent on the number of points used to evaluate the kernel before the integrations. From the tests, 50 points seem to
yield a good tradeoff between speed and acuracy, but if only the monopole is used, 10 points seem to be enough.
Building
To install, do
python -m pip install git+https://github.com/dforero0896/geofptax.git --U
By default, the package is installed without the C extension, to do so (if you want to test or use the C module instead) do
GEOFPTAX_CEXT=1 python -m pip install git+https://github.com/dforero0896/geofptax.git --U
Contents:
- API Reference
bk_multip()bk_sugiyama_multip()bkeff_r_scalar()bkeff_r_vmap()bkeff_sugiyama()bkeff_sugiyama_vec()compute_basis_grid()compute_sugiyama_multipoles()cosab()ext_bk_mp()f2_ker()g2_ker()g2_ker_vec()geo_fac()geo_fac_damped()geo_fac_sugiyama_simple()geo_fac_sugiyama_simple_vec()geo_fac_vec()integrate_2d_gauss()integrate_bkeff_r()interpol_ker()jax_leggauss()legendre_recurrence()pt_kernel()pt_pk_1loop()vec_integrate_bkeff_r()weights_trapz()z1_ker()z1_ker_vec()z2_ker()z2_ker_vec()