Cross-ILC methods paper

Paper:

Cross-ILC technique:
- For kSZ: Compute cross-spectrum between tSZ-free and CIB-free ILC maps to measure the kSZ power spectrum.
- For CMB lensing: Pass tSZ-free and CIB-free ILC maps in the two legs of lensing quadratic estimator to mitigate both tSZ and CIB-induced lensing biases.

make_paper_plots.ipynb: Reproduce plots in the paper.
Fisher formalism:
- s1_get_fisher_matrix.ipynb: Forecasts total SNR of the kSZ power spectrum and also cosmological / reionisation constraints for current and future CMB surveys.
- s2_analyse_fisher_matrix.ipynb: Analyse the results from s1_get_fisher_matrix.ipynb. This notebook will help in making tables and figures.
get_bandpower_errors.ipynb: Obtain the bandpower errors for different ILC techniques / CMB experiments analytically using Knox formula.
estimate_bias_using_fisher.ipynb: Estimate biases in kSZ and cosmological parameters due to unmodelled CIB/tSZ residuals. This used a modifed Fisher formalism based on Amara & Refregier 2007.
- There are also couple of more example notebooks to understand Fisher-based bias estimation in fisher_bias folder.
  - example1_bias_in_line_fitting.ipynb - Bias estimation in simple line fitting and a comparison of the result using a lileihood approach.
  - example2_bias_in_cosmo.ipynb - Inject biases as xx $\sigma$ of cosmological parameters and recover them using the modified Fisher appraoch.
get_radio_residuals.ipynb: Modelling the power spectrum of radio sources and also get the ILCed radio residuals analytically.

Data:
- lagache_2019_ns150_radio.dat: Radio $dN/dS$ = Lagache et al. 2020 model (arXiv: 1911.09466) used to model radio point source power.
- ilc_weights_residuals_agora_fg_model.npy: ILC residuals for multiple ILC combinations and all experiments considered in this work.
- CAMB: CMB power spectra and derivatives of the power spectra as a function of cosmological parameters:
  - Both Planck 2018 and Planck 2015 are available.
  - Planck 2018 is the default.
  - These are used for Fisher forecasting.
  - Both lensed and unlensed versions are available.
  - Fiducial: Lensed spectra.
ILC results:
- ilc_weights_residuals_agora_fg_model.npy - ILC residuals and weights for AGORA foreground model.
Fisher results:
- fiducial - Results from TTEETE; 3mJy source mask; Total parameters = 14: 6 LCDM + 2 radio parameters + 1 CIB/tSZ + 1 kSZ amplitude + 2 homogeneous kSZ + 2 reionisation kSZ.
Lensing results:
- Lensing N0: Results available for standard QE, MH18, and the cross-ILC.
  - Simply do: ls *qewithcmbmv* or ls *mh* or ls *xilc* to see the files in the linked Lensing N0 folder.
- Lensing cross-correlations: