lib.samples.mesh.Mesh¶

class lib.samples.mesh.Mesh(mesh, dims, edges=None, nodes=None, isdensity=True)¶

Bases: cosmopipe.lib.utils.BaseClass

Class that represents a mesh. TODO: decide what to do with MPI… scatter mesh? Gather mesh?

Initialize Mesh.

Parameters

mesh (array) – Mesh array.
dims (list) – List of labels (one for each mesh dimension).
edges (list, default=None) – List of edges (one array for each mesh dimension). If None, use nodes instead. Mesh shape must be length of edges - 1 along each dimension.
nodes (list, default=None) – List of mesh nodes (one array for each mesh dimension).
isdensity (bool, default=True) – Whether to normalize binned (weighted) samples by mesh volume (True).

Methods

`argmax`	Return coordinates of maximum along axes `axes` or dimensions `dims`.
`argmean`	Return mean coordinates along axes `axes` or dimensions `dims`.
`copy`	Return shallow copy of `self`.
`deepcopy`	Return deep copy.
`from_samples`	Initialize mesh from samples.
`from_state`	Instantiate and initalize class with state dictionary.
`get_axes`	Return axes (axes) corresponding to dimensions `dims`.
`get_levels`	Return density threshold(s) corresponding to confidence level(s) `targets`.
`get_sigmas`	Return density thresholds corresponding to n-sigmas confidence levels `sigmas`.
`integrate`	Integrate mesh along axes `axes` or dimensions `dims`.
`interpolate`	Interpolate mesh at provided points.
`is_mpi_broadcast`
`is_mpi_gathered`
`is_mpi_root`
`is_mpi_scattered`
`load`	Load class in numpy binary format from disk.
`load_auto`	If different formats are possible, this method should between them based on file name extension.
`log_critical`
`log_debug`
`log_error`
`log_info`
`log_warning`
`max`	Return mesh maximum value.
`mnodes`	Return meshgrid of nodes along axes `axes` or dimensions `dims`.
`ranges`	Return mesh extension along axes `axes` or dimensions `dims`.
`save`	Save class to disk.
`save_auto`	If different formats are possible, this method should between them based on file name extension.
`save_txt`	Save mesh as ASCII file.
`volume`	Return volume array along axes `axes` or dimensions `dims`.
`widths`	Return cell sizes along axes `axes` or dimensions `dims`.

Attributes

`logger`
`mpiattrs`	MPI attributes
`mpicomm`
`mpiroot`
`mpistate`
`ndim`	Return number of dimensions.
`shape`	Return mesh shape.

__add__(other)¶: Addition of two meshes.

__call__(axes=None)¶: Return mesh reduced to axes axes or dimensions dims. Mesh is integrated along other dimensions.

argmax(axes=None)¶: Return coordinates of maximum along axes axes or dimensions dims.

argmean(axes=None)¶: Return mean coordinates along axes axes or dimensions dims.

copy()¶: Return shallow copy of self.

deepcopy()¶: Return deep copy.

classmethod from_samples(samples, weights=None, dims=None, bins=30, method='cic', bw_method='scott')¶

Initialize mesh from samples. TODO: scattered samples?

Parameters

samples (array) – Array of positions, with coordinates along first axes.
weights (array, default=None) – Sample weights, defaults to 1.
dims (list) – List of labels (one for each mesh dimension).
bins (int, default=30) – Number of bins i.e. mesh nodes to use along each dimension.
method (string) –
Method to interpolate (weighted) samples on mesh, either:
- ’gaussian_kde’: Gaussian kernel density estimation, based on scipy.stats.gaussian_kde
- ’cic’ : Cloud-in-Cell assignment
- ’histo’ : simple binning.
bw_method (string, default='scott') – If method is 'gaussian_kde', method to determine KDE bandwidth, see scipy.stats.gaussian_kde.

Returns

mesh

Return type

Mesh

classmethod from_state(state, mpiroot=0, mpicomm=None)¶: Instantiate and initalize class with state dictionary.

get_axes(dims=None)¶: Return axes (axes) corresponding to dimensions dims.

get_levels(targets)¶: Return density threshold(s) corresponding to confidence level(s) targets.

get_sigmas(sigmas)¶

Return density thresholds corresponding to n-sigmas confidence levels sigmas.

Parameters: sigmas (int, list, array) – Sigmas to get density thresholds for. If scalar, return thresholds up to (including) sigmas.
Returns: levels – Density thresholds.
Return type: array

integrate(axes=None)¶: Integrate mesh along axes axes or dimensions dims.

interpolate(points, method='linear')¶

Interpolate mesh at provided points.

Parameters

points (array) – Array of points, with coordinates along last axis.
method (string) – Interpolation method, ‘linear’ or ‘nearest’ (see scipy.interpolate.interpn())

Returns

values – Mesh values at interpolated points.

Return type

array

classmethod load(filename, mpiroot=0, mpicomm=None)¶: Load class in numpy binary format from disk. If the loaded state contains __class__ and that exists in cls._registry, return instance of cls._registry[__class__] (instead of cls).

load_auto(*args, **kwargs)¶: If different formats are possible, this method should between them based on file name extension.

max()¶: Return mesh maximum value.

mnodes(axes=None)¶: Return meshgrid of nodes along axes axes or dimensions dims.

property mpiattrs¶: MPI attributes

property ndim¶: Return number of dimensions.

ranges(axes=None)¶: Return mesh extension along axes axes or dimensions dims.

save(filename)¶: Save class to disk.

save_auto(*args, **kwargs)¶: If different formats are possible, this method should between them based on file name extension.

save_txt(filename, header='', fmt='%.8e', delimiter=' ', **kwargs)¶

Save mesh as ASCII file. TODO: needs major improvement, e.g. what about dimension names?

Parameters

filename (string) – File name where to save mesh.
kwargs (dict) – Arguments for numpy.savetxt().

property shape¶: Return mesh shape.

volume(axes=None)¶: Return volume array along axes axes or dimensions dims.

widths(axes=None)¶: Return cell sizes along axes axes or dimensions dims.