BaryonForge.Runners.Map2DRunner module

class BaryonForge.Runners.Map2DRunner.DefaultRunnerGrid(HaloNDCatalog, GriddedMap, epsilon_max, model, use_ellipticity=False, mass_def=MassDef(Delta=200, rho_type=critical), include_pixel_size=True, verbose=True)

Bases: object

A utility class for handling input/output operations related to halo ND catalogs and gridded maps.

The DefaultRunnerGrid class provides methods to manage and process data associated with halo ND catalogs, including constructing rotation matrices and generating coordinate arrays. It supports operations in both 2D and 3D contexts and handles optional ellipticity-based calculations.

Parameters:

• HaloNDCatalog (object) – An instance of a HaloNDCatalog, containing data about halos and their properties. It must have a cosmology attribute to specify the cosmological parameters.

• GriddedMap (object) – An instance representing a GriddedMap, either 2D or 3D, where halo information will be mapped.

• epsilon_max (float) – A parameter specifying the maximum size, in units of halo radius, of cutouts made around each halo during painting/baryonification.

• model (object, optional) – An object that generates profiles or displacements. For example, see Baryonification2D or Pressure

• use_ellipticity (bool, optional) – A flag indicating whether to use ellipticity in calculations. Default is False. If set to True, a NotImplementedError is raised, as this mode has not yet been implemented for curved, full-sky baryonification.

• mass_def (object, optional) – An instance of a mass definition object from the CCL (Core Cosmology Library), specifying the mass definition to be used. Default is ccl.halos.massdef.MassDef(200, ‘critical’).

• verbose (bool, optional) – A flag to enable verbose output for logging or debugging purposes. Default is True.

HaloNDCatalog

The HaloNDCatalog instance.

Type:

object

GriddedMap

The GriddedMap instance.

Type:

object

cosmo

The cosmology object extracted from HaloNDCatalog.

Type:

object

model

The model used for baryonification or profile painting.

Type:

object

epsilon_max

The maximum radius, in halo radius units, of cutouts around halos.

Type:

float

mass_def

The mass definition object.

Type:

object

verbose

Whether verbose output is enabled. Defaults to True.

Type:

bool, optional

include_pixel_size

Only used when painting, not baryonifying. If True, then the returned map is multiplied by the area/volume of the pixel. Thus, painting with a density profile results in a Mass map. Defaults to True.

Type:

bool, optional

use_ellipticity

Whether to use ellipticity in calculations. Defaults to False.

Type:

bool, optional

build_Rmat(A, q)

Constructs a rotation matrix based on the input vector A and ellipticity parameter q.

coord_array(*args)

Flattens and stacks input arrays into a 2D array of coordinates.

Raises:

• AssertionError – If use_ellipticity is True and required columns (‘q_ell’, ‘c_ell’, ‘A_ell’) are missing in the HaloNDCatalog.

• NotImplementedError – If attempting to use the 3D ellipticity method, which is not yet verified.

build_Rmat(A, q)

Constructs a rotation matrix based on the input vector and ellipticity parameter.

This method normalizes the input vector A and calculates the rotation matrix using the ellipticity parameter q. For 2D vectors, it uses the shear transformation. For 3D vectors, a not yet verified method is provided but raises a NotImplementedError.

Parameters:

• A (ndarray) – A 1D array representing the vector to be rotated. It will be normalized within the method.

• q (float) – The ellipticity parameter, used to compute the shear transformation.

Returns:

Rmat – A 2x2 or 3x3 rotation matrix, depending on the dimensionality of the input vector.

Return type:

ndarray

Raises:

• ValueError – If the input vector A is 1-dimensional.

• NotImplementedError – If a 3D rotation is attempted, indicating that the method is not yet verified for 3D vectors.

coord_array(*args)

Flattens and stacks input arrays into a 2D array of coordinates.

This method takes multiple input arrays, flattens each, and stacks them column-wise to create a single 2D array where each row represents a coordinate.

Parameters:

*args (list of ndarrays) – Arrays to be flattened and stacked. Each input array represents one dimension of the coordinates. All arrays must have the same shape.

Returns:

coords – A 2D array of shape (N, M) where N is the total number of elements (after flattening) and M is the number of input arrays. Each row represents a coordinate.

Return type:

ndarray

class BaryonForge.Runners.Map2DRunner.BaryonifyGrid(HaloNDCatalog, GriddedMap, epsilon_max, model, use_ellipticity=False, mass_def=MassDef(Delta=200, rho_type=critical), include_pixel_size=True, verbose=True)

Bases: DefaultRunnerGrid

A class to apply baryonification to a gridded map using a halo catalog.

The BaryonifyGrid class inherits from DefaultRunnerGrid and is designed to process a gridded map by applying baryonification techniques to adjust the matter distribution. It uses a halo catalog to determine the necessary adjustments based on halo properties, cosmological parameters, and a specified model.

The inputted grid should be MASS grid rather than density grid. This is because the method uses pix = 0 to identify empty pixels.

process()

Processes the gridded map by applying baryonification and returns the modified grid.

pick_indices(center, width, Npix)

Helper method that selects and returns indices around a center point, accounting for periodic boundary conditions.

pick_indices(center, width, Npix)

Selects and returns indices around a center point, accounting for periodic boundary conditions.

This method takes a central index and a width and returns an array of indices around the center, wrapping around if the indices go beyond the boundaries of the grid. This is used to get cutouts around a given halo.

Parameters:

• center (int) – The central index around which indices are selected.

• width (int) – The half-width of the selection range. The method selects indices from center - width to center + width.

• Npix (int) – The total number of pixels along one dimension of the grid. Used to wrap indices for periodic boundary conditions.

Returns:

inds – An array of selected indices, wrapped around the boundaries if necessary.

Return type:

ndarray

process()

Applies baryonification to the gridded map using the halo catalog.

This method iterates over each halo in the HaloNDCatalog, calculating the necessary displacements based on the halo’s mass, position, and other properties. It uses the given model to compute the displacement, updates the gridded map accordingly, and ensures that the total mass remains conserved.

Returns:

new_map – A 2D or 3D numpy array representing the modified grid after baryonification.

Return type:

ndarray

Raises:

• AssertionError – If the sum of the new map values does not match the sum of the original map values, indicating an error in pixel regridding.

• NotImplementedError – If the 3D ellipticity method is attempted, which is currently not supported.

Notes

• This method supports both 2D and 3D gridded maps.

• The ParamTabulatedProfile model is required if property keys are used in the model.

• Non-finite displacement values are set to zero to avoid issues with map updates.

class BaryonForge.Runners.Map2DRunner.PaintProfilesGrid(HaloNDCatalog, GriddedMap, epsilon_max, model, use_ellipticity=False, mass_def=MassDef(Delta=200, rho_type=critical), include_pixel_size=True, verbose=True)

Bases: DefaultRunnerGrid

A class to paint profiles onto a gridded map using a halo catalog.

The PaintProfilesGrid class inherits from DefaultRunnerGrid and is designed to generated a grid of a given property (mass, temperature, pressure) by painting halo profiles. It uses a halo catalog to determine the necessary profiles based on halo properties, cosmological parameters, and a specified model.

The returned map by default includes an integration over pixel size. For example, passing model = density_profile will return a map of the Mass = density * dV and not the density alone.

process()

Processes the gridded map by painting baryonic profiles and returns the modified grid.

pick_indices(center, width, Npix)

Helper function that Selects and returns indices around a center point, accounting for periodic boundary conditions.

pick_indices(center, width, Npix)

Selects and returns indices around a center point, accounting for periodic boundary conditions.

This method takes a central index and a width and returns an array of indices around the center, wrapping around if the indices go beyond the boundaries of the grid.

Parameters:

• center (int) – The central index around which indices are selected.

• width (int) – The half-width of the selection range. The method selects indices from center - width to center + width.

• Npix (int) – The total number of pixels along one dimension of the grid. Used to wrap indices for periodic boundary conditions.

Returns:

inds – An array of selected indices, wrapped around the boundaries if necessary.

Return type:

ndarray

process()

Applies profile painting to the gridded map using the halo catalog.

This method iterates over each halo in the HaloNDCatalog, calculating the profile contributions based on the halo’s mass, position, and other properties. It uses the provided model to compute the profile and updates the gridded map accordingly.

Returns:

new_map – A 2D or 3D numpy array representing the modified grid after painting the baryonic profiles.

Return type:

ndarray

Raises:

• AssertionError – If a model is not provided or if the provided model is not an instance of ParamTabulatedProfile when property keys are used.

• ValueError – If use_ellipticity is True and the 3D map painting method is attempted, which is currently not supported.

Notes

• This method supports both 2D and 3D gridded maps.

• The ParamTabulatedProfile model is required if property keys are used in the model.

• Non-finite profile values are set to zero to avoid issues with map updates.

class BaryonForge.Runners.Map2DRunner.PaintProfilesAnisGrid(HaloNDCatalog, GriddedMap, epsilon_max, model, Tracer_model, Mtot_model, background_val, global_tracer_fraction, mass_def=MassDef(Delta=200, rho_type=critical), include_pixel_size=True, use_ellipticity=False, verbose=True)

Bases: PaintProfilesGrid

process()

Applies profile painting to the gridded map using the halo catalog.

This method iterates over each halo in the HaloNDCatalog, calculating the profile contributions based on the halo’s mass, position, and other properties. It uses the provided model to compute the profile and updates the gridded map accordingly.

Returns:

new_map – A 2D or 3D numpy array representing the modified grid after painting the baryonic profiles.

Return type:

ndarray

Raises:

• AssertionError – If a model is not provided or if the provided model is not an instance of ParamTabulatedProfile when property keys are used.

• ValueError – If use_ellipticity is True and the 3D map painting method is attempted, which is currently not supported.

Notes

• This method supports both 2D and 3D gridded maps.

• The ParamTabulatedProfile model is required if property keys are used in the model.

• Non-finite profile values are set to zero to avoid issues with map updates.

BaryonForge.Runners.Map2DRunner.regrid_pixels_2D(grid, pix_positions, pix_values)

Redistributes pixel values onto a 2D (regular, square) grid considering periodic boundary conditions.

This function takes a list of pixel positions and their associated values, then redistributes these values onto a specified 2D grid. It accounts for overlap and periodic boundary conditions, ensuring proper handling of edge cases where offsets are significantly larger than the grid size.

Parameters:

• grid (ndarray) – A 2D numpy array representing the grid onto which pixel values will be redistributed. The grid is modified in place. Must be a square grid.

• pix_positions (ndarray of shape (N, 2)) – An array of pixel positions, where each position is given by (x, y) coordinates. These coordinates specify where the displaced pixels are located.

• pix_values (ndarray of shape (N,)) – An array of pixel values corresponding to each position in pix_positions. These values are redistributed across the grid based on the pixel’s overlap with the grid.

Notes

• The function uses Numba’s @njit decorator for just-in-time compilation, optimizing performance.

• Periodic boundary conditions are handled explicitly to ensure proper wrapping around the grid edges.

• This function assumes that both grid and pix_positions use a zero-based index system and that the grid is square with shape (N, N).

BaryonForge.Runners.Map2DRunner.regrid_pixels_3D(grid, pix_positions, pix_values)

Redistributes pixel values onto a 3D grid considering periodic boundary conditions.

This function takes a list of 3D pixel positions and their associated values, then redistributes these values onto a specified 3D grid. It accounts for overlap and periodic boundary conditions, ensuring proper handling of edge cases where offsets are significantly larger than the grid size.

Parameters:

• grid (ndarray) – A 3D numpy array representing the grid onto which pixel values will be redistributed. The grid is modified in place. Must be a cubic grid.

• pix_positions (ndarray of shape (N, 3)) – An array of pixel positions, where each position is given by (x, y, z) coordinates. These coordinates specify where the displaced pixels are located.

• pix_values (ndarray of shape (N,)) – An array of pixel values corresponding to each position in pix_positions. These values are redistributed across the grid based on overlap.

Notes

• The function uses Numba’s @njit decorator for just-in-time compilation, optimizing performance.

• Periodic boundary conditions are handled explicitly to ensure proper wrapping around the grid edges.

• This function assumes that both grid and pix_positions use a zero-based index system and that the grid is cubic with shape (N, N, N).