flint.predict.crystalball

Items in and out the model visibility prediction using the crystalball python package.

Classes

CrystalBallOptions

Options related to running crystal ball

Functions

cli(→ None)

crystalball_predict(→ flint.ms.MS)

A very simply wrapper around the Crystalball.predict function. Basic

get_parser(→ argparse.ArgumentParser)

Module Contents

class flint.predict.crystalball.CrystalBallOptions(/, **data: Any)[source]

Bases: flint.options.BaseOptions

Options related to running crystal ball

crystallball_wsclean_pol_mode: list[str] = ['i'][source]

The polarisation of the wsclean model that was generated

memory_fraction: float = 0.75[source]

The fraction of available memory to use to define the target chunk size

model_chunks: int = 0[source]

Number of sky model components that are processed in a single chunk. If 0 it will be set automatically. Default is 0.

row_chunks: int = 0[source]

Number of rows of input MS that are processed in a single chunk. If 0 it will be set automatically. Default is 0.

flint.predict.crystalball.cli() → None[source]
flint.predict.crystalball.crystalball_predict(ms: flint.ms.MS, crystalball_options: CrystalBallOptions, wsclean_source_list_path: pathlib.Path | None = None, dask_client: dask.distributed.Client | None = None, output_column: str = 'MODEL_DATA', update_crystalball_options: dict[str, Any] | None = None) → flint.ms.MS[source]

A very simply wrapper around the Crystalball.predict function. Basic checks to ensure that the BB6 style source model path exists, which is the format used by the wsclean -save-source-list option.

If no wsclean_source_list_path is specified one is guess from the name of the input MS.path.

Parameters:

  • ms (MS) – The MS instance whose path is to the measurement set to predict into.

  • crystalball_options (CrystalBallOptions) – Options that control the crystalball.predict call.

  • wsclean_source_list_path (Path | None, optional) – The path to the file with the model components to predict. If None an attempt is made to find it from the MS. Defaults to None.

  • dask_client (Client | None, optional) – A specialised Dask distributed task. If None one is created by crystalball. Defaults to None.

  • output_column (str, optional) – The column to predict into. The MS.model_column will reflect this. Defaults to “MODEL_DATA”.

  • update_crystalball_options (dict[str, Any] | None, optional) – Update options to the provided crystalball_options. Defaults to None.

Returns:

The MS that was predicted into, with the model_column set appropriately.

Return type:

MS

flint.predict.crystalball.get_parser() → argparse.ArgumentParser[source]