Quick Start Guide

This guide will get you started with CosmoForge’s main packages quickly.

CosmoForge 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 estimation.

  • PICSLike (picslike) — pixel-space likelihood.

  • CosmoForge (cosmoforge) — umbrella metapackage; depends on the three above.

Getting Started with CosmoCore

Parameter Management

CosmoCore uses a centralized parameter system:

from cosmocore.settings import InputParams

# Create with defaults
params = InputParams()

# Check key parameters
print(f"HEALPix nside: {params.nside}")
print(f"Maximum l: {params.lmax}")
print(f"Field labels: {params.labels}")
print(f"Number of pixels: {params.npix}")

Configuration Files

Use YAML files for configuration:

# config.yaml
nside: 32
lmax: 128
labels: ["T", "E", "B"]          # Analysis field labels
physical_labels: ["T", "Q", "U"] # Map field labels (or "TQU", "T1_T2")
input_convention: Dl              # "Cl" (default) or "Dl" for input files
fwhmarcmin: 5.0
apply_pixwin: true

# Load configuration
params = InputParams.read_parameter_file('config.yaml')

Field Expansion

CosmoForge supports automatic field label expansion:

# Concatenated single-character fields
physical_labels: ["QU"]      # Expands to ["Q", "U"]
labels: ["TEB"]              # Expands to ["T", "E", "B"]

# Underscore-separated multi-character fields
labels: ["T1_T2"]            # Expands to ["T1", "T2"]
physical_labels: ["LOW_HIGH"] # Expands to ["LOW", "HIGH"]

# Mixed configurations
labels: ["T", "QU", "E1_E2"]  # Expands to ["T", "Q", "U", "E1", "E2"]

Mathematical Utilities

Access optimized mathematical functions:

from cosmocore.basics import legendre_00, legendre_22, legendre_02

x = 0.5     # cos(theta)
lmax = 10

P_l_TT = legendre_00(x, lmax)   # spin-0
P_l_EE = legendre_22(x, lmax)   # spin-2 auto
P_l_TE = legendre_02(x, lmax)   # spin-0 x spin-2 cross

Getting Started with QUBE

Fisher Matrix Analysis

from qube import Fisher

# Initialize Fisher analysis
fisher = Fisher("config/fisher_config.yaml")
fisher.run()

if fisher.rank == 0:
    errors = fisher.get_error_bars()

QML Power Spectrum Estimation

from qube import Spectra

# Initialize QML analysis
spectra = Spectra("config/qml_config.yaml")
spectra.run()

if spectra.rank == 0:
    # Default: deconvolved estimates of true C_ell
    cl = spectra.get_power_spectra()
    # Other modes: "decorrelated", "convolved"
    cl_decorr = spectra.get_power_spectra(mode="decorrelated")

Next Steps