CosmoForge Documentation

CosmoForge is a comprehensive Python framework for power spectrum estimation and likelihood analysis of spin-0 and spin-2 fields on the sphere, using Fisher matrix, Quadratic Maximum Likelihood (QML), and pixel-based likelihood methods. While widely applicable to any sky signal (e.g. CMB, galaxy surveys, 21 cm), it is particularly optimized for the analysis of partial-sky, noisy observations with complex noise covariance.

Architecture

CosmoForge is a uv workspace of three runtime packages plus an umbrella metapackage:

  • CosmoCore (cosmocore): shared algebra, fields, computation bases, I/O.

  • QUBE (qube, PyPI name qube-qml): Fisher and QML power-spectrum estimation.

  • PICSLike (picslike): Pixel-based Inference with Correlated-Skies Likelihood — pixel-space likelihood analysis.

  • CosmoForge (cosmoforge): umbrella metapackage; one-line install for the whole framework.

Key Features

  • Fisher Matrix Analysis: Fast parameter forecasting and covariance estimation

  • QML Power Spectrum Estimation: Optimal power spectrum recovery from noisy data

  • Pixel-Based Likelihood: Direct likelihood evaluation in map pixel space

  • High-Performance Computing: Numba-optimized functions and MPI parallelization support

  • HEALPix Integration: Full support for HEALPix pixelization schemes

  • Flexible Field Management: Support for scalar (spin-0) and tensor (spin-2) fields

  • Instrumental Effects: Comprehensive beam and noise modeling

Quick Start

# Fisher Matrix Analysis
from qube import Fisher
fisher = Fisher(params_file="config/fisher_config.yaml")
fisher.run()

# Pixel-Based Likelihood
from picslike import PICSLike
picslike = PICSLike(params_file="config/pixel_config.yaml")
picslike.run()

# Core mathematical utilities
from cosmocore import InputParams
params = InputParams()
print(f"HEALPix resolution: nside={params.nside}")

Contents

Documentation:

API Reference:

Development: