nifits.backend package

Submodules

nifits.backend.backend module

Backend module

This module handles basic computation based on the NIFITS standard for representing the instrument. It reconstructs the model-in-a-kit following the standard.

It offers a limited capability to be powered with computing backends that have a numpy-compatible API.

class nifits.backend.backend.MovingCollection(series)[source]

Bases: object

property coords
property coords_rad
property coords_radial

Returns the radial coordinates of points. (rho, theta) ([unit], [rad]).

property coords_shaped
series: PointCollection
class nifits.backend.backend.NI_Backend(nifits=None, module=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Bases: object

A class to produce calculations based on the NIFITS standard.

add_instrument_definition(nifits_instrument=None, force=False, verbose=True)[source]

Adds the instrument definition to the model.

Parameters:

  • nifits_instrument (nifits) – NIFITS object

  • force (bool) – Bool if True, then the existing extensions will be replaced

  • verbose (bool) – Get more printed information

add_observation_data(nifits_data=None, force=False, verbose=True)[source]

Adds the observation data to the model.

Parameters:

  • nifits_instrument – NIFITS object

  • force (bool) – Bool if True, then the existing extensions will be replaced

  • verbose (bool) – Get more printed information

create_fov_function_all(md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Constructs the method to get the chromatic phasor given by injection for all the time series.

Parameters:

md – A module for the computations

Sets up self.ni_fov.xy2phasor

downsample(Is)[source]

Downsample flux from the NI_OSWAVELENGTH bins to OI_WAVELENGTH bins. Expected shape is : (n_frames, n_wl, n_outputs, n_points), the method simply applies the NI_DSAMP matrix along the second axis (1).

Parameters:

Is – ArrayLike [flux] input the result computed with the oversampled wavelength channels. (n_frames, n_wlold, n_outputs, n_points)

Returns:

ArrayLike [flux] (n_frames, n_wlnew, n_outputs, n_points)

Return type:

Ids

Returns

geometric_phasor(alpha, beta, include_mod=True, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Returns the complex phasor corresponding to the locations of the family of sources

Parameters:

  • alpha – The coordinate matched to X in the array geometry

  • beta – The coordinate matched to Y in the array geometry

  • anarray – The array geometry (n_input, 2)

  • include_mod – Include the modulation phasor

Returns:

A vector of complex phasors

get_Is(xs, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Get intensity from an array of sources.

get_KIs(Iarray, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Get the prost-processed observable from an array of output intensities. The post-processing matrix K is taken from self.nifits.ni_kmat.K

Parameters:

  • I – (n_frames, n_wl, n_outputs, n_batch)

  • md (ModuleType) – a python module with a numpy-like interface.

Returns:

The vector \(\boldsymbol{\kappa} = \mathbf{K}\cdot\mathbf{I}\)

get_all_outs(alphas, betas, kernels=False, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Compute the transmission map for an array of coordinates. The map can be seen as equivalent collecting power expressed in [m^2] for each point sampled so as to facilitate comparison with models in Jansky multiplied by the exposure time of each frame (available in nifits.ni_mod.int_time).

Parameters:

  • alphas – ArrayLike [rad] 1D array of coordinates in right ascension

  • betas – ArrayLike [rad] 1D array of coordinates in declination

  • kernels – (bool) if True, then computes the post-processed observables as per the KMAT matrix.

Returns:

the raw transmission output. if kernels is True: the differential observable.

Return type:

if kernels is False

Hint

Shape: (n_frames, n_wl, n_outputs, n_points)

get_modulation_phasor(md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Computes and returns the modulation phasor [n_wl, n_input]

The modulation phasor includes is computed in units of collected amplitude so that the output of the transmission map in square modulus provides equivalent collecting power in m^2 for each point of the map. This is done to facilitate the usag of the map with models in jansky.

get_moving_outs(alphas, betas, kernels=False, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Compute the transmission map for an array of coordinates. The map can be seen as equivalent collecting power expressed in [m^2] for each point sampled so as to facilitate comparison with models in Jansky multiplied by the exposure time of each frame (available in nifits.ni_mod.int_time).

Parameters:

  • alphas – ArrayLike [rad] 1D array of coordinates in right ascension

  • betas – ArrayLike [rad] 1D array of coordinates in declination

  • kernels – (bool) if True, then computes the post-processed observables as per the KMAT matrix.

Returns:

the raw transmission output. if kernels is True: the differential observable.

Return type:

if kernels is False

Hint

Shape: (n_frames, n_wl, n_outputs, n_points)

moving_geometric_phasor(alphas, betas, include_mod=True, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Returns the complex phasor corresponding to the locations of the family of sources

Parameters:

  • alpha : (n_frames, n_points) The coordinate matched to X in the array geometry

  • beta : (n_frames, n_points) The coordinate matched to Y in the array geometry

  • anarray : The array geometry (n_input, 2)

  • include_mod : Include the modulation phasor

Returns : A vector of complex phasors

nifits

Backend object. Typically created empty. The nifits instrument information are added later with:

  • .add_instrument_definition

  • .add_observation_data

Parameters:

  • nifits – NIFITSClass

  • module – A backend module for advanced math.

plot_recorded(cmap='viridis', outputs=None, nrows_ncols=None, res_x=1000, res_y=500, interp='nearest')[source]
class nifits.backend.backend.PointCollection(aa=None, bb=None, unit=None)[source]

Bases: object

A class to hold arrays of coordinates. Handy to compute

the transmission map on a large number of points.

Units default to mas.

Parameters:

  • aa (Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – [unit (mas)] (ArrayLike) first coordinate flat array, typically RA.

  • bb (Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – [unit (mas)] (ArrayLike) second coordinate flat array, typically Dec.

Constructors:

  • from_uniform_disk :

  • from_grid :

  • from_centered_square:

  • from_segment :

Modificators:

  • __add__ : basically a concatenation

  • transform : Linear transformation in 3D by a matrix

Handles:

  • coords : The array values as first provided

  • coords_rad : The array values, converted from self.unit into radians.

  • coords_quantity: Return the values as a quantity.

  • coords_radial : Returns the radial coordinates (rho,theta)

  • extent : The [left, right, top, bottom] extent (used for some plots).

aa: Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]] = None
bb: Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]] = None
property coords

Returns a tuple with the alpha and beta coordinates in flat arrays.

property coords_rad
property coords_radial

Returns the radial coordinates of points. (rho, theta) ([unit], [rad]).

property coords_shaped
classmethod from_centered_square_grid(radius, resolution, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Create a centered square cartesian grid object

Parameters:

  • radius – The radial extent of the grid.

  • resolution – The number of pixels across the width.

classmethod from_grid(a_coords, b_coords, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>, unit=Unit("mas"))[source]

Create a point collection as a cartesian grid.

Parameters:

  • a_coords (Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – The array of samples along the first axis (typically alpha)

  • b_coords (Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – The array of samples along the second axis (typically beta, the second dimension)

  • md (ModuleType) – Module of choice for the computation

  • unit (Unit) – Units for a_coords and b_coords

Handles:

classmethod from_segment(start_coords, end_coords, n_samples, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>, unit=Unit("mas"))[source]

Create a point collection as a cartesian grid.

Parameters:

  • start_coords (Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – The (a,b) array of the starting point. (typically alpha, beta)

  • end_coords (Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – The (a,b) array of the ending point. (typically alpha, beta)

  • n_sameples – The number of samples along the line.

Handles:

classmethod from_uniform_disk(radius=None, n=10, phi_0=0.0, offset=array([0., 0.]), md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>, unit=Unit("mas"))[source]

Create a point collection as a uniformly sampled disk.

Parameters:

  • radius (float) – [mas] The angular radius of the disk to model

  • n (int) – The total number of points to create

  • phi_0 (float) – [rad] Arbitrary angle to initialiaze pattern

  • offset (Union[Buffer, _SupportsArray[dtype[Any]], _NestedSequence[_SupportsArray[dtype[Any]]], bool, int, float, complex, str, bytes, _NestedSequence[Union[bool, int, float, complex, str, bytes]]]) – [mas] An offset of the disk location with respect to the center of the field of view

  • md (ModuleType) – The numerical module to use as backend

  • unit (Unit) – The unit to use for interactions

plot_frame(z=None, frame_index=0, wl_index=0, out_index=0, mycmap=None, marksize_increase=1.0, colorbar=True, xlabel=True, title=True)[source]
transform(matrix, md=<module 'numpy' from '/lhome/romain/miniconda3/envs/aiitec-jax/lib/python3.12/site-packages/numpy/__init__.py'>)[source]

Produce a linear transform of the coordinates.

Parameters:

  • matrix – A transformation matrix (3D)

  • md – A module to do the operation

unit: Unit = None
nifits.backend.backend.col_row_numbers(n, col_ceiling=4)[source]

Module contents