ConvertReferenceCatalogTask

class lsst.meas.algorithms.ConvertReferenceCatalogTask(*, output_dir=None, **kwargs)

Bases: Task

Class for producing HTM-indexed reference catalogs from external catalog data.

This implements an indexing scheme based on hierarchical triangular mesh (HTM). The term index really means breaking the catalog into localized chunks called shards. In this case each shard contains the entries from the catalog in a single HTM trixel

For producing catalogs this task makes the following assumptions about the input catalogs:

• RA, Dec are in decimal degrees.

• Epoch is available in a column, in a format supported by astropy.time.Time.

• There are either no off-diagonal covariance terms, or there are all the five-dimensional covariance terms (between RA, Dec, proper motion, and parallax). In the latter case, a custom ConvertRefcatManager must exist to handle the covariance terms.

Parameters

output_dirstr

The path to write the output files to, in a subdirectory defined by DatasetConfig.ref_dataset_name.

Attributes Summary

canMultiprocess

Methods Summary

makeSchema(dtype)	Make the schema to use in constructing the persisted catalogs.
run(inputFiles)	Index a set of files comprising a reference catalog.

Attributes Documentation

canMultiprocess = False

Methods Documentation

makeSchema(dtype)

Make the schema to use in constructing the persisted catalogs.

Parameters

dtypenumpy.dtype

Data type describing each entry in config.extra_col_names for the catalogs being ingested.

Returns

schemaAndKeyMaptuple of (lsst.afw.table.Schema, dict)

A tuple containing two items: - The schema for the output source catalog. - A map of catalog keys to use in filling the record

run(inputFiles)

Index a set of files comprising a reference catalog.

Outputs are persisted in the butler repository.

Parameters

inputFileslist

A list of file paths to read.