Foreground fitting (ps_eor.fgfit)

Classical foreground fitting and subtraction methods.

All fitters implement the same small interface: call run(data_cube, noise_cube) and receive a FitterResult containing the foreground model in FitterResult.fit and the residual data in FitterResult.sub. Both outputs retain the input cube geometry, so the residual can be passed directly to flagging or power-spectrum estimation.

A common deterministic fit models each visibility with a smooth function of frequency. The noise cube supplies the per-frequency noise scale used by the fit:

from ps_eor import fgfit

fitter = fgfit.PolyForegroundFit(
    deg=3,
    fit_type='power_poly',
)
result = fitter.run(data_cube, noise_cube)

foreground = result.fit
residual = result.sub

Valid polynomial families are 'poly', 'bernstein', 'power_poly', and 'power_bernstein'. The power variants are often useful for smooth foreground spectra.

For a component-based separation, PcaForegroundFit fits the real and imaginary data independently and also exposes its component representation:

result = fgfit.PcaForegroundFit(n_cmpt=3).run(data_cube, noise_cube)
first_component = result.inverse_transform(0)

These fitters provide inexpensive deterministic baselines. Probabilistic foreground separation and posterior sampling are available through ps_eor.ml_gpr.

class ps_eor.fgfit.FitterResult(cube_fit, cube_sub)

Bases: object

Foreground model and residual cubes.

fit

Fitted foreground model.

sub

Input data minus the model.

class ps_eor.fgfit.MixForegroundResult(cube_fit, cube_sub, cube_mix, inv_trans)

Bases: FitterResult

A fitted model with its component mixing representation.

inverse_transform(mode)

Reconstruct the sky from the mixing model.

Parameters:

mode – component index to reconstruct alone, or None for all.

Returns:

the reconstructed cube.

Return type:

CartDataCube

get_component(n)

The n-th fitted component as a one-channel cube.

Returns:

component n of the mixing cube.

Return type:

CartDataCube

class ps_eor.fgfit.AbstractForegroundFitter

Bases: object

Interface implemented by foreground fitters.

run(data_cube, data_cube_noise)

Fit the foregrounds in data_cube and separate them from the data.

Parameters:

• data_cube – the data to fit (a CartDataCube).

• data_cube_noise – a noise cube of the same geometry, used to weight the fit and set thresholds.

Returns:

the foreground model (.fit) and residual (.sub).

Return type:

FitterResult

class ps_eor.fgfit.NoAction

Bases: AbstractForegroundFitter

A no-op fitter: a zero model, leaving the data untouched.

run(i_cube, v_cube)

Return a zero foreground model, keeping i_cube as the residual.

Returns:

.fit is zeros, .sub is i_cube unchanged.

Return type:

FitterResult

class ps_eor.fgfit.GmcaForegroundFit(n_cmpt, mints=0, do_wave_transform=False, do_poly_fit=0)

Bases: AbstractForegroundFitter

Foreground separation with Generalized Morphological Component Analysis.

run(data_cube, data_cube_noise)

Separate foregrounds with GMCA (see AbstractForegroundFitter.run()).

Returns:

the GMCA model (.fit) and residual (.sub).

Return type:

FitterResult

class ps_eor.fgfit.PcaForegroundFit(n_cmpt)

Bases: AbstractForegroundFitter

Foreground separation using independent PCA fits to real and imaginary data.

run(data_cube, data_cube_noise)

Separate foregrounds with PCA (see AbstractForegroundFitter.run()).

Returns:

the model (.fit), residual (.sub) and the component mixing cube (.mix / MixForegroundResult.get_component()).

Return type:

MixForegroundResult

class ps_eor.fgfit.PolyForegroundFit(deg, fit_type)

Bases: AbstractForegroundFitter

Fit a smooth spectral model independently to each visibility.

run(data_cube, data_cube_noise)

Fit a smooth spectral model per visibility (see AbstractForegroundFitter.run()).

Returns:

the smooth model (.fit) and residual (.sub).

Return type:

FitterResult