Schemas

Definitions not specific to a particular dataset schema.

Quantities

Quantity DataArrays encode general-purpose data that relates to certain physical quantities. They are typically associated with an SI unit.

class QuantityInSecondsArray

Bases: object

Quantity with units of seconds

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInSecondsArray
Attributes:
units			's'	Default: None
type			'quantity'	Default: 'quantity'

class QuantityInHertzArray

Bases: object

Quantity with units of Hertz

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInHertzArray
Attributes:
units			'Hz'	Default: None
type			'quantity'	Default: 'quantity'

class QuantityInMetersArray

Bases: object

Quantity with units of meters

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInMetersArray
Attributes:
units			'm'	Default: None
type			'quantity'	Default: 'quantity'

class QuantityInMetersPerSecondArray

Bases: object

Quantity with units of meters per second

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInMetersPerSecondArray
Attributes:
units			'm/s'	Default: None
type			'quantity'	Default: 'quantity'

class QuantityInRadiansArray

Bases: object

Quantity with units of radians

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInRadiansArray
Attributes:
units			'rad'	Default: None
type			'quantity'	Default: 'quantity'

class QuantityInKelvinArray

Bases: object

Quantity with units of Kelvins

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInKelvinArray
Attributes:
units			'K'	Default: None
type			'quantity'	Default: 'quantity'

class QuantityInKelvinPerJanskyArray

Bases: object

Quantity with units of K/Jy (sensitivity in gain curve)

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInKelvinPerJanskyArray
Attributes:
units			'K/Jy'	Default: None
type			'quantity'	Default: 'quantity'

class QuantityInPascalArray

Bases: object

Quantity with units of Pa

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInPascalArray
Attributes:
units			'Pa'	Default: None
type			'quantity'	Default: 'quantity'

class QuantityInPerSquareMetersArray

Bases: object

Quantity with units of /m^2

	Dimensions	Dtype	Model	Description
data	[]	<f8	QuantityInPerSquareMetersArray
Attributes:
units			'/m^2'	Default: None
type			'quantity'	Default: 'quantity'

Measures

As with python-casacore measures, measures are quantities that are interpreted in relation to a specified reference frame (such as UTC for TimeArray or FK5 for SkyCoordArray). Measure definitions are aligned with astropy coordinate naming conventions as much as possible. The table below outlines the different types of XRADIO measures:

class TimeArray

Representation of a time measure.

astropy.time.Time serves as the reference implementation. Data can be converted as follows:

astropy.time.Time(data * astropy.units.Unit(attrs['units'][0]),
                  format=attrs['format'], scale=attrs['scale'])

All formats that express time as floating point numbers since an epoch are permissible, so at present the realistic options are:

• mjd (from 1858-11-17 00:00:00 UTC)

• unix (from 1970-01-01 00:00:00 UTC)

• unix_tai (from 1970-01-01 00:00:00 TAI)

• cxcsec (from 1998-01-01 00:00:00 TT)

• gps (from 1980-01-06 00:00:00 UTC)

	Dimensions	Dtype	Model	Description
data	[]	<f8	TimeArray	Time since epoch, typically in seconds (see units).
Attributes:
type			'time'	Array type. Should be "time". Default: 'time'
units			's'	Units to associate with axis Default: 's'
scale			'tai', 'tcb', 'tcg', 'tdb', 'tt', 'ut1' or 'utc'	Time scale of data. Must be one of ('tai', 'tcb', 'tcg', 'tdb', 'tt', 'ut1', 'utc'), see astropy.time.Time Default: 'utc'
format			'unix', 'mjd', 'cxcsec' or 'gps'	Time representation and epoch, see TimeArray. Default: 'unix'

class SpectralCoordArray

Measures array for data variables and attributes that are spectral coordinates.

	Dimensions	Dtype	Model	Description
data	[]	<f8	SpectralCoordArray
Attributes:
units			'Hz'	Default: 'Hz'
observer			'REST', 'BARY', 'TOPO', 'gcrs', 'icrs', 'hcrs', 'lsrk', 'lsrd' or 'lsr'	Capitalized reference observers are from casacore. TOPO implies creating astropy earth_location. Astropy velocity reference frames are lowercase. Note that Astropy does not use the name ‘TOPO’ (telescope centric) rather it assumes if no velocity frame is given that this is the default. When converting from MSv2 and casacore frequency frames, the following translations from casacore to astropy frame names are applied: GEO=>gcrs, LSRK=>lsrk, LSRD=>lsrd Default: 'icrs'
type			'spectral_coord'	Default: 'spectral_coord'

class SkyCoordArray

Measures array for data variables that are sky coordinates, used in FieldSourceXds

	Dimensions	Dtype	Model	Description
data	[sky_dir_label] or [sky_dis_label]	<f8	SkyCoordArray
Attributes:
units			'm' or 'rad'	Default: None
type			'sky_coord'	Default: 'sky_coord'
frame			'icrs', 'fk5', 'fk4', 'fk4noterms', 'galactic', 'galactocentric', 'supergalactic', 'altaz', 'hadec', 'gcrs', 'cirs', 'itrs', 'hcrs', 'teme', 'tete', 'precessedgeocentric', 'geocentricmeanecliptic', 'barycentricmeanecliptic', 'heliocentricmeanecliptic', 'geocentrictrueecliptic', 'barycentrictrueecliptic', 'heliocentrictrueecliptic', 'heliocentriceclipticiau76', 'custombarycentricecliptic', 'lsr', 'lsrk', 'lsrd' or 'galacticlsr'	Possible values are astropy.coordinates.SkyCoord frames. Several casacore frames found in MSv2 are translated to astropy frames as follows: * AZELGEO => altaz * J2000 => fk5 * ICRS => icrs From fixvis docs: clean and the im tool ignore the reference frame claimed by the UVW column (it is often mislabelled as ITRF when it is really FK5 or J2000) and instead assume the (u, v, w)s are in the same frame as the phase tracking center. calcuvw does not yet force the UVW column and field centers to use the same reference frame! Default: 'icrs'

class LocationArray

Measure type used for example in antenna_xds/ANTENNA_POSITION, weather_xds/STATION_POSITION, field_and_source_xds(ephemeris)/OBSERVER_POSITION.

Data dimensions can be CartesianPosLabel or EllipsoidDirLabel or EllipsoidDisLabel

	Dimensions	Dtype	Model	Description
data	[ellipsoid_dir_label] or [ellipsoid_dis_label] or [cartesian_pos_label]	<f8	LocationArray
Attributes:
units			'm' or 'rad'	Units of the location coordinates (typically ‘m’ or ‘rad’). Default: None
frame			'ITRS' or 'Undefined'	Reference frame. Can be ITRS (assumed for all Earth locations) or Undefined (used in non-Earth locations). Default: None
coordinate_system			'geocentric', 'planetcentric', 'geodetic', 'planetodetic', 'orbital' or 'topocentric'	Can be geocentric/planetcentric, geodetic/planetodetic, orbital Default: None
origin_object_name			str	earth/sun/moon/etc. Default: None
(ellipsoid)			'GRS80', 'WGS84' or 'WGS72'	Ellipsoid used in geodetic Earth locations (with EllipsoidDirLabel and EllipsoidDirLabel coordinate) Default: None
type			'location'	Measure type. Should be "location". Default: 'location'

class DopplerArray

Doppler measure information for the frequency coordinate

	Dimensions	Dtype	Model	Description
data	[]	<f8	DopplerArray
Attributes:
type			'doppler'	Coordinate type. Should be "doppler". Default: 'doppler'
units			'ratio' or 'm/s'	Units to associate with axis, [ratio]/[m/s] Default: 'm/s'
doppler_type			'radio', 'optical', 'z', 'ratio', 'true', 'relativistic', 'beta' or 'gamma'	Allowable values: radio, optical, z, ratio, true, relativistic, beta, gamma. Astropy only has radio and optical. Using casacore types: https://casadocs.readthedocs.io/en/stable/notebooks/memo-series.html?highlight=Spectral%20Frames#Spectral-Frames Default: 'radio'

Coordinate Labels

For some types of measures, the data consists of values that are labeled using coordinate labels. These labels provide context for interpreting the data:

Coordinate Label Name	Values	Related Measures Type
ellipsoid_dir_label	lon, lat	location
ellipsoid_dis_label	height	location
cartesian_pos_label	x, y, z	location
cartesian_local_pos_label	p, q, r	location
galactic_sky_dir_label	lon, lat	sky_coord
local_sky_dir_label	az, alt	sky_coord
local_sky_dis_label	dist	sky_coord
sky_dir_label	ra, dec	sky_coord
sky_dist_label	dist	sky_coord
uvw_label	u, v, w	uvw
receptor_label	pol_0, pol_1	quantity
tone_label	tone_0, tone_1	spectral_coord

Measures Example

The following example illustrates how measures information is included in both a data variable (FIELD_PHASE_CENTER) and a coordinate (time). The FIELD_PHASE_CENTER data variable has the dimensions time and sky_dir_label. Note that the sky_coord measure requires only the sky_dir_label dimension, not the time dimension.

import xarray as xr
phase_center = xr.DataArray()

import numpy as np
import xarray as xr
import pandas as pd

# Create an empty Xarray Dataset.
xds = xr.Dataset()

# Create the time coordinate with time measures attributes.
time = xr.DataArray(pd.date_range('2000-01-01', periods=3).astype('datetime64[s]').astype(int), dims='time', attrs={'type': 'time', 'units': 's', 'format':'unix', 'scale':'utc'})

# Create FIELD_PHASE_CENTER data variable with coordinates time x sky_dir_label.
coords = {'time': time,
          'sky_dir_label': ['ra', 'dec']}

data = np.array([[-2.10546176, -0.29611873],
       [-2.10521098, -0.29617315],
       [-2.1050196, -0.2961987]])

xds['FIELD_PHASE_CENTER'] = xr.DataArray(data, coords=coords, dims=['time', 'sky_dir_label'])

# Add sky_coord measures attributes to FIELD_PHASE_CENTER.
xds['FIELD_PHASE_CENTER'].attrs = {
    "type": "sky_coord",
    "units": "rad",
    "frame": "icrs",
}

xds

<xarray.Dataset> Size: 96B
Dimensions:             (time: 3, sky_dir_label: 2)
Coordinates:
  * time                (time) int64 24B 946684800 946771200 946857600
  * sky_dir_label       (sky_dir_label) <U3 24B 'ra' 'dec'
Data variables:
    FIELD_PHASE_CENTER  (time, sky_dir_label) float64 48B -2.105 ... -0.2962

xarray.Dataset

• Dimensions:
  • time: 3
  • sky_dir_label: 2
• Coordinates: (2)
  • time
    (time)
    int64
    946684800 946771200 946857600

    type :

    time

    units :

    s

    format :

    unix

    scale :

    utc

    array([946684800, 946771200, 946857600])

  • sky_dir_label
    (sky_dir_label)
    <U3
    'ra' 'dec'

    array(['ra', 'dec'], dtype='<U3')

• Data variables: (1)
  • FIELD_PHASE_CENTER
    (time, sky_dir_label)
    float64
    -2.105 -0.2961 ... -2.105 -0.2962

    type :

    sky_coord

    units :

    'rad'

    frame :

    icrs

    array([[-2.10546176, -0.29611873],
           [-2.10521098, -0.29617315],
           [-2.1050196 , -0.2961987 ]])

• Indexes: (2)
  • time
    PandasIndex

    PandasIndex(Index([946684800, 946771200, 946857600], dtype='int64', name='time'))

  • sky_dir_label
    PandasIndex

    PandasIndex(Index(['ra', 'dec'], dtype='object', name='sky_dir_label'))

• Attributes: (0)

# Example of creating an Astropy SkyCoord object from the FIELD_PHASE_CENTER data variable.
from astropy.coordinates import SkyCoord
astropy_skycoord = SkyCoord(ra=xds.FIELD_PHASE_CENTER.sel(sky_dir_label='ra').values,dec=xds.FIELD_PHASE_CENTER.sel(sky_dir_label='dec').values,unit='rad',frame=xds.FIELD_PHASE_CENTER.attrs['frame'])
astropy_skycoord

<SkyCoord (ICRS): (ra, dec) in deg
    [(239.36592723, -16.96635346), (239.38029586, -16.9694715 ),
     (239.39126113, -16.97093541)]>