cmbroom 0.2.1

A Python package for blind component separation of microwave sky maps.

BROOM: Blind Reconstruction Of signals from Observations in the Microwaves

BROOM is a Python package for blind component separation and Cosmic Microwave Background (CMB) data analysis.

---

📦 Installation

You can install the base package using:

pip install cmbroom

This installs the core functionality.
If you plan to use the few functions that depend on pymaster, you must install it separately (version >=2.4).

---

🔧 To include pymaster automatically:

You can install cmbroom along with its optional pymaster dependency by running:

pip install cmbroom[pymaster]

However, pymaster requires some additional system libraries to be installed before running the above command.

✅ On Ubuntu/Debian:

sudo apt update
sudo apt install build-essential python3-dev libfftw3-dev libcfitsio-dev libgsl-dev

✅ On macOS (using Homebrew):

brew install fftw cfitsio gsl

Documentation

A detailed introduction to the parameters and simulation pipeline is available in:

• tutorials/tutorial_satellite.ipynb
• configs/config_demo.yaml — Example configuration file

Component separation methods are covered in:

• tutorials/tutorial_satellite.ipynb
• tutorials/tutorial_satellite_part2.ipynb

Power spectrum estimation is demonstrated in:

• tutorials/tutorial_spectra.ipynb

For partial-sky, ground-based experiment analysis, see:

• tutorials/tutorial_groundbased.ipynb

🔗 Full online documentation:
👉 https://alecarones.github.io/broom/

References

Paper on broom package is in preparation.

If you use the following methodologies please cite the corresponding papers:

• ILC or NILC: Bennett et al., 2003, Delabrouille et al., 2009
• cMILC: Remazeilles et al., 2021, Carones et al., 2024
• MC-ILC or MC-NILC: Carones et al., 2023
• PILC: Fernández-Cobos et al., 2016
• cPILC: Adak, 2021
• GILC, GNILC, GPILC: Remazeilles et al., 2011, Planck Collaboration, 2016
• foreground diagnostic: Carones et al., 2024
• power spectrum computation: Gorski et al., 2005, Zonca et al., 2019, Alonso et al., 2019

📦 Dependencies

This package relies on several scientific Python libraries:

• astropy>=6.0.1
• numpy>1.18.5
• scipy>=1.8
• healpy>=1.15
• pysm3>=3.3.2
• mtneedlet>=0.0.5