torchplite is a PyTorch implementation of the Planck 2018 Lite likelihood for the Cosmic Microwave Background (CMB) power spectra. This package provides a convenient and efficient way to compute the log-likelihood of CMB power spectra given a cosmological model.
You can install torchplite via pip from PyPI:
pip install torchplite
This package requires Python 3.6 or later and PyTorch 1.7 or later.
To use the torchplite package, you can import the PlanckLitePy class and create an instance with the desired settings:
from torchplite import PlanckLitePy
# Initialize the PlanckLitePy object
planck = PlanckLitePy(year=2018, spectra="TTTEEE", use_low_ell_bins=False)
# Load the power spectra
import numpy as np
ls, Dltt, Dlte, Dlee = np.genfromtxt("path/to/your/data/Dl_planck2015fit.dat", unpack=True)
# Compute the log-likelihood
ellmin = int(ls[0])
loglikelihood = planck.loglike(Dltt, Dlte, Dlee, ellmin)You can customize the behavior of the PlanckLitePy object by changing its constructor parameters:
year: The Planck data release year (2015 or 2018).spectra: The CMB power spectra to use ("TTTEEE" for TT, TE, and EE or "TT" for TT only).use_low_ell_bins: Whether to include low-ell bins in the likelihood calculation (True or False).
To run the tests, you can use the unittest module:
python -m unittest discover tests
this will run all the test cases defined in the tests directory.
This project is under the MIT License. See the LICENSE file for more details.
This is a PyTorch implementation of the planck-lite-py code by Heather Prince.