FMM.pipeline module

class FMM.pipeline.PipelineEnd2End(comm)

Bases: object

FMM Pipeline.

Wrapper for End-2-End pipeline. It added class one after the others by running method.run().

Methods

main

main(specific_file=None)

class FMM.pipeline.PipelineFrequencyMapMaking(comm, file, params)

Bases: object

Instance to reconstruct frequency maps using QUBIC abilities.

Parameters:

comm

MPI communicator

filestr

used to create the forlder for data saving

paramsdict

dictionary containing all the simulation parameters

Methods

get_H()	Acquisition operator
get_averaged_nus()	Average frequency
get_components_fgb()	Components FGbuster
get_convolution()	QUBIC resolutions.
get_dict([key])	QUBIC dictionary.
get_input_map()	Input maps.
get_preconditioner()	PCG Preconditioner.
get_random_value()	Random value
get_sky_config()	Sky configuration.
get_tod([noise])	Simulated TOD.
pcg(d, x0, seenpix)	Preconditioned Conjugate Gradiant algorithm.
run()	Run the FMM Pipeline.

get_H()

Acquisition operator

Method to compute QUBIC acquisition operators.

get_averaged_nus()

Average frequency

Method to average QUBIC frequencies according to the number of reconstructed frequency maps.

Returns:

nus_ave – array containing the averaged QUBIC frequencies.

Return type:

array_like

get_components_fgb()

Components FGbuster

Method to build a dictionary containing all the wanted components to generate sky maps. Based on FGBuster.

Returns:

dict_comps – Dictionary containing the component instances.

Return type:

dict

get_convolution()

QUBIC resolutions.

Method to define expected QUBIC angular resolutions (radians) as function of frequencies.

Returns:

• fwhm_in (array_like) – Intrinsic resolutions, used to build the simulated TOD.

• fwhm_out (array_like) – Output resolutions. If we don’t apply convolutions during reconstruction, array of zeros.

• fwhm_rec (array_like) – Reconstructed resolutions. Egal the output resolutions if we apply convolutions during reconstructions, evaluate through analytic formula otherwise.

get_dict(key='in')

QUBIC dictionary.

Method to modify the qubic dictionary.

Parameters:

key (str, optional) – Can be “in” or “out”. It is used to build respectively the instances to generate the TODs or to reconstruct the sky maps, by default “in”.

Returns:

dict_qubic – Modified QUBIC dictionary.

Return type:

dict

get_input_map()

Input maps.

Function to get the input maps from PySM3.

Returns:

maps_in – Input maps \((N_{rec}, 12 \times N^{2}_{side}, N_{stk})\).

Return type:

array_like

get_preconditioner()

PCG Preconditioner.

Computed using the formula: To be added.

Returns:

M – Preconditioner for PCG algorithm.

Return type:

DiagonalOperator

get_random_value()

Random value

Method to build a random seed.

Returns:

seed – Random seed.

Return type:

int

get_sky_config()

Sky configuration.

Method that read ‘params.yml’ file and create dictionary containing sky emission model.

Returns:

dict_sky – Sky config dictionary.

Return type:

dict

Notes

Note that the key denote the emission and the value denote the sky model using PySM convention. For CMB, seed denote the realization.

Example

d = {‘cmb’:seed, ‘dust’:’d0’, ‘synchrotron’:’s0’}

get_tod(noise=False)

Simulated TOD.

Method that compute observed TODs with \(\vec{TOD} = H \cdot \vec{s} + \vec{n}\), with H the QUBIC operator, \(\vec{s}\) the sky signal and \(\vec{n}\) the instrumental noise`.

Parameters:

noise (bool, optional) – True if you want to simulate noise in your data, False otherwise, by default False.

Returns:

TOD – Simulated TOD \((N_{rec}, 12 \times N^{2}_{side}, N_{stk})\).

Return type:

array_like

pcg(d, x0, seenpix)

Preconditioned Conjugate Gradiant algorithm.

Solve the map-making equation iteratively : \((H^T . N^{-1} . H) . x = H^T . N^{-1} . d\).

The PCG used for the minimization is intrinsequely parallelized (e.g see PyOperators).

Parameters:

• d (array_like) – Array containing the TODs generated previously \((N_{rec}, 12 \times N^{2}_{side}, N_{stk})\).

• x0 (array_like) – Starting point of the PCG algorithm \((N_{rec}, 12 \times N^{2}_{side}, N_{stk})\).

• seenpix (array_like) – Boolean array to define the pixels seen by QUBIC \((N_{rec}, 12 \times N^{2}_{side}, N_{stk})\).

Returns:

solution – Reconstructed maps \((N_{rec}, 12 \times N^{2}_{side}, N_{stk})\).

Return type:

array_like

run()

Run the FMM Pipeline.

Method to run the whole pipeline from TOD generation from sky reconstruction by reading params.yml file.