yaw.catalogs.treecorr.TreecorrCatalog#

class yaw.catalogs.treecorr.TreecorrCatalog(data: DataFrame, ra_name: str, dec_name: str, *, patch_name: str | None = None, patch_centers: BaseCatalog | Coordinate | None = None, n_patches: int | None = None, redshift_name: str | None = None, weight_name: str | None = None, cache_directory: str | None = None, progress: bool = False)[source]#

Bases: BaseCatalog

An implementation of the BaseCatalog using TreeCorr for the pair counting.

Note

The current implementation is not very efficient, because the internal fields of the underlying treecorr.Catalog must be rebuilt every time the catalog is iterated in redshift bins for the pair counting.

Warning

Currently this backend does not support restoration from cache and raises an NotImplementedError` when calling the from_cache() method.

Build a catalogue from in-memory data.

Catalogs should be instantiated through the factory class, see yaw.catalogs.NewCatalog.from_dataframe().

Methods

`__init__`(data, ra_name, dec_name, *[, ...])	Build a catalogue from in-memory data.
`correlate`(config, binned[, other, linkage, ...])	Count pairs between objects at a given separation and in bins of redshift.
`from_cache`(cache_directory[, progress])
`from_file`(filepath, patches, ra, dec, *[, ...])	Build a catalogue from data file.
`from_treecorr`(cat)	Create a new instace from a `treecorr.Catalog`.
`get_max_redshift`()	Get the maximum redshift or `None` if not available.
`get_min_redshift`()	Get the minimum redshift or `None` if not available.
`get_totals`()	Get an array of the sum of weights or number of objects in each patch.
`has_redshifts`()	Indicates whether the `redshifts()` attribute holds data.
`has_weights`()	Indicates whether the `weights()` attribute holds data.
`is_loaded`()	Indicates whether the catalog data is loaded.
`iter_bins`(z_bins[, allow_no_redshift])	Iterate the catalogue in bins of redshift.
`load`()	Permanently load data from cache into memory.
`to_treecorr`()	Get the internal `treecorr.Catalog` instance.
`true_redshifts`(config[, sampling_config, ...])	Compute a histogram of the object redshifts from the binning defined in the provided configuration.
`unload`()	Unload data from memory if a disk cache is provided.

Attributes

`centers`	Get a vector of sky coordinates of the patch centers in radians.
`dec`	Get an array of the declination values in radians.
`ids`	Return a list of unique patch indices in the catalog.
`n_patches`	The number of spatial patches of this catalogue.
`patch`	Get the patch indices of each object as array.
`pos`	Get a vector of the object sky positions in radians.
`ra`	Get an array of the right ascension values in radians.
`radii`	Get a vector of angular separations in radians that describe the patch sizes.
`redshifts`	Get the redshifts as array or `None` if not available.
`total`	Get the sum of weights or the number of objects if weights are not available.
`weights`	Get the object weights as array or `None` if not available.

property centers: CoordSky#

Get a vector of sky coordinates of the patch centers in radians.

Returns:: yaw.core.coordinates.CoordSky

correlate(config: Configuration, binned: bool, other: TreecorrCatalog = None, linkage: PatchLinkage | None = None, progress: bool = False) → NormalisedCounts | dict[str, NormalisedCounts][source]#

Count pairs between objects at a given separation and in bins of redshift.

If another catalog instance is passed to other, then pairs are formed between these catalogues (cross), otherwise pairs are formed with the catalog (auto). Pairs are counted in bins of redshift, as defined in the configuration object (config). Pairs are only considered within fixed angular scales that are computed from the physical scales in the configuration and the mid of the current redshift bin.

Parameters:

config (yaw.Configuration) – Configuration object that defines measurement scales, redshift binning, cosmological model, and various backend specific parameters.
binned (bool) – Whether to apply the redshift binning to the second catalogue (see other).
other (Catalog instance, optional) – Second catalog instance used for cross-catalogue pair counting. Catalogue must use the same backend.
linkage (PatchLinkage, optional) – Linkage object that defines with patches must be correlated for a given scales and which patch combinations can be skipped. Can be used for the scipy backend to count pairs consistently between multiple catalogue instances.
progress (bool) – Show a progress indication, depends on backend.

There are three different modes of operation that are determined by the combination of the binned and other parameters:

If no second catalogue is provided, pairs are counted within the catalogue while applying the redshift binning.
If a second catalogue is provided and binned=True, pairs are counted between the catalogues and the binning is applied to both cataluges.
If a second catalogue is provided and binned=False, the redshift binning is not applied to the second catalogue, otherwise above.

The catalogue from the calling instance of correlate() has always redshift binning applied.

property dec: ndarray[Any, dtype[float64]]#: Get an array of the declination values in radians.

classmethod from_cache(cache_directory: str, progress: bool = False) → NoReturn[source]#

Raises:: NotImplementedError –

Warning

Currently this backend does not support restoration from cache.

Build a catalogue from data file.

Catalogs should be instantiated through the factory class, see yaw.catalogs.NewCatalog.from_file().

classmethod from_treecorr(cat: Catalog) → TreecorrCatalog[source]#: Create a new instace from a treecorr.Catalog.

get_max_redshift() → float[source]#: Get the maximum redshift or None if not available.

get_min_redshift() → float[source]#: Get the minimum redshift or None if not available.

get_totals() → ndarray[Any, dtype[float64]][source]#: Get an array of the sum of weights or number of objects in each patch.

has_redshifts() → bool[source]#: Indicates whether the redshifts() attribute holds data.

has_weights() → bool[source]#: Indicates whether the weights() attribute holds data.

property ids: list[int]#: Return a list of unique patch indices in the catalog.

is_loaded() → bool[source]#

Indicates whether the catalog data is loaded.

Always True if no cache is used. If the catalog is unloaded, data will be read from cache every time data is accessed.

iter_bins(z_bins: NDArray[np.float64], allow_no_redshift: bool = False) → Iterator[tuple[Interval, TreecorrCatalog | EmptyCatalog]][source]#

Iterate the catalogue in bins of redshift.

Parameters:

z_bins (NDArray) – Edges of the redshift bins.
allow_no_redshift (bool) – If true and the data has no redshifts, the iterator yields the whole catalogue at each iteration step.

Yields:

(tuple) –

tuple containing:

intv (pandas.Interval): the selection for this bin.
cat (TreecorrCatalog): instance containing the data for this bin.

load() → None[source]#

Permanently load data from cache into memory.

Raises a CachingError if no cache is configured.

property n_patches: int#: The number of spatial patches of this catalogue.

property patch: ndarray[Any, dtype[int64]]#: Get the patch indices of each object as array.

property pos: CoordSky#

Get a vector of the object sky positions in radians.

Returns:: yaw.core.coordinates.CoordSky

property ra: ndarray[Any, dtype[float64]]#: Get an array of the right ascension values in radians.

property radii: DistSky#

Get a vector of angular separations in radians that describe the patch sizes.

The radius of the patch is defined as the maximum angular distance of any object from the patch center.

Returns:: yaw.core.coordinates.DistSky

property redshifts: ndarray[Any, dtype[float64]] | None#: Get the redshifts as array or None if not available.

to_treecorr() → Catalog[source]#: Get the internal treecorr.Catalog instance.

property total: float#: Get the sum of weights or the number of objects if weights are not available.

true_redshifts(config: Configuration, sampling_config: ResamplingConfig | None = None, progress: bool = False) → HistData[source]#

Compute a histogram of the object redshifts from the binning defined in the provided configuration.

Parameters:

config (Configuration) – Defines the bin edges used for the histogram.
sampling_config (ResamplingConfig, optional) – Specifies the spatial resampling for error estimates.
progress (bool) – Show a progress bar.

Returns:

Object holding the redshift histogram

Return type:

HistData

unload() → None[source]#: Unload data from memory if a disk cache is provided.

property weights: ndarray[Any, dtype[float64]]#: Get the object weights as array or None if not available.