sirius._sirius_utils._cngi_io

ms_selection(infile, outfile=None, verbose=False, spw=None, field=None, times=None, baseline=None, scan=None, scanintent=None, array=None, uvdist=None, observation=None, polarization=None)[source]
read_generic_table(infile, subtables=False, timecols=None, ignore=None)[source]

read generic casacore table format to xarray dataset loaded in memory

Parameters

  • infile (str) – Input table filename. To read a subtable simply append the subtable folder name under the main table (ie infile = ‘/path/mytable.tbl/mysubtable’)

  • subtables (bool) – Whether or not to include subtables underneath the specified table. If true, an attribute called subtables will be added to the returned xds. Default False

  • timecols (list) – list of column names to convert to numpy datetime format. Default None leaves times as their original casacore format.

  • ignore (list) – list of column names to ignore and not try to read. Default None reads all columns

Returns

Return type

xarray.core.dataset.Dataset

read_ms(infile, rowmap=None, subtables=False, expand=False, chunks=(22000, 512, 2))[source]

Read legacy format MS to xarray Visibility Dataset

The MS is partitioned by DDI, which guarantees a fixed data shape per partition. This results in separate xarray dataset (xds) partitions contained within a main xds (mxds).

Parameters

  • infile (str) – Input MS filename

  • rowmap (dict) – Dictionary of DDI to tuple of (row indices, channel indices). Returned by ms_selection function. Default None ignores selections

  • subtables (bool) – Also read and include subtables along with main table selection. Default False will omit subtables (faster)

  • expand (bool) – Whether or not to return the original flat row structure of the MS (False) or expand the rows to time x baseline dimensions (True). Expanding the rows allows for easier indexing and parallelization across time and baseline dimensions, at the cost of some conversion time. Default False

  • chunks (4-D tuple of ints) – Shape of desired chunking in the form of (time, baseline, channel, polarization). Larger values reduce the number of chunks and speed up the reads at the cost of more memory. Chunk size is the product of the four numbers. Default is (400, 400, 64, 2). None disables re-chunking and returns native chunk size from table row reads

Returns

Main xarray dataset of datasets for this visibility set

Return type

xarray.core.dataset.Dataset

write_generic_table(xds, outfile, subtable='', cols=None, verbose=False)[source]

Write generic xds contents back to casacore table format on disk

Parameters

  • xds (xarray.Dataset) – Source xarray dataset data

  • outfile (str) – Destination filename (or parent main table if writing subtable)

  • subtable (str) – Name of the subtable being written, triggers special logic to add subtable to parent table. Default ‘’ for normal generic writes

  • cols (str or list) – List of cols to write. Default None writes all columns

write_main_table_slice(xda, outfile, ddi, col, full_shape, starts)[source]

Write an xds row chunk to the corresponding main table slice

write_ms(mxds, outfile, infile=None, subtables=False, modcols=None, verbose=False, execute=True)[source]

Write ms format xds contents back to casacore table format on disk

Parameters

  • mxds (xarray.Dataset) – Source multi-xarray dataset (originally created by read_ms)

  • outfile (str) – Destination filename

  • infile (str) – Source filename to copy subtables from. Generally faster than reading/writing through mxds via the subtables parameter. Default None does not copy subtables to output.

  • subtables (bool) – Also write subtables from mxds. Default of False only writes mxds attributes that begin with xdsN to the MS main table. Setting to True will write all other mxds attributes to subtables of the main table. This is probably going to be SLOW! Use infile instead whenever possible.

  • modcols (list) – List of strings indicating what column(s) were modified (aka xds data_vars). Different logic can be applied to speed up processing when a data_var has not been modified from the input. Default None assumes everything has been modified (SLOW)

  • verbose (bool) – Whether or not to print output progress. Since writes will typically execute the DAG, if something is going to go wrong, it will be here. Default False

  • execute (bool) – Whether or not to actually execute the DAG, or just return it with write steps appended. Default True will execute it

visplot(xda, axis=None, overplot=False, drawplot=True, tsize=250)[source]

Plot a preview of Visibility xarray DataArray contents

Parameters

  • xda (xarray.core.dataarray.DataArray) – input DataArray to plot

  • axis (str or list or xarray.core.dataarray.DataArray) – Coordinate(s) within the xarray DataArray, or a second xarray DataArray to plot against. Default None uses range. All other coordinates will be maxed across dims

  • overplot (bool) – Overlay new plot on to existing window. Default of False makes a new window for each plot

  • drawplot (bool) – Display plot window. Should pretty much always be True unless you want to overlay things in a Jupyter notebook.

  • tsize (int) – target size of the preview plot (might be smaller). Default is 250 points per axis

Returns

Return type

Open matplotlib window

read_image(infile, masks=True, history=True, chunks=(1000, 1000, 1, 4), verbose=False)[source]

Read casacore format Image to xarray Image Dataset format

Parameters

  • infile (str) – Input image filename (.image or .fits format)

  • masks (bool) – Also read image masks as additional image data_vars. Default is True

  • history (bool) – Also read history log table. Default is True

  • chunks (4-D tuple of ints) – Shape of desired chunking in the form of (l, m, chan, pol). Default is (1000, 1000, 1, 4) Note: chunk size is the product of the four numbers (up to the actual size of the dimension)

Returns

new xarray Datasets of Image data contents

Return type

xarray.core.dataset.Dataset

write_image_slice(xda, outfile, col, starts)[source]

Write image xda chunk to the corresponding image table slice

write_image(xds, outfile, portion='IMAGE', masks=True, history=True, verbose=False, execute=True)[source]

Read casacore format Image to xarray Image Dataset format

Parameters

  • xds (xarray.Dataset) – Image xarray dataset to write

  • outfile (str) – Output image filename (.image format)

  • portion (str) – Name of the data_var in the xds that corresponds to the image data. Default ‘IMAGE’

  • masks (bool) – Also write the masks to the output. Can be used instead of infile parameter. Default True

  • history (bool) – Also write the history log file to the output. Can be used instead of infile paramter. Default True

  • verbose (bool) – Whether or not to print output progress. Since writes will typically execute the DAG, if something is going to go wrong, it will be here. Default False

  • execute (bool) – Whether or not to actually execute the DAG, or just return it with write steps appended. Default True will execute it