.. _ApplyBeam.json#/:

ApplyBeam
=========

Apply the LOFAR beam model, or the inverse of it.


.. _ApplyBeam.json#/inputs:

inputs
******


.. _ApplyBeam.json#/inputs/type:

type
++++

:type: ``string``

:default: ``applybeam``

:doc: Case-insensitive step type; must be 'applybeam' `.`


.. _ApplyBeam.json#/inputs/beammode:

beammode
++++++++

:type: ``enum``

:default: ``default``

:doc: Beam mode to apply, can be ``array_factor``, ``element``, ``full``, or ``default`` (same as ``full``). The mode ``full`` applies both the element beam and the array factor `.`

:symbols: ``array_factor``, ``element``, ``full``, ``default``


.. _ApplyBeam.json#/inputs/beamproximitylimit:

beamproximitylimit
++++++++++++++++++

:type: ``float``

:default: ``60``

:doc: Specified in arcseconds, and if non-zero, sources that are near each other are clustered and the beam is only calculated for each cluster. This trades an increase in speed for a very low level of loss in accuracy. This parameter is ignored in the ApplyBeam step itself, and should instead be used as part of predict, ddecal, gaincal, h5parmpredict, etc. This parameter has been introduced in Dec 2020 `.`


.. _ApplyBeam.json#/inputs/direction:

direction
+++++++++

:type: ``string?``

:doc: A RA/Dec value specifying in what direction to correct the beam. See ``phaseshift.phasecenter`` for syntax. If empty, the beam is corrected in the direction of the current phase center.


.. _ApplyBeam.json#/inputs/invert:

invert
++++++

:type: ``boolean?``

:default: ``True``

:doc: Invert the corrections, to correct the data. Default is ``true``. If you want to corrupt the data, set it to ``false`` `.`


.. _ApplyBeam.json#/inputs/onebeamperpatch:

onebeamperpatch
+++++++++++++++

:type: ``string?``

:default: ``False``

:doc: Deprecated: use beamproximitylimit instead. When ``true``, computes the beam only for the center of each patch (saves computation time, but you should set this to ``false`` for large patches). In the ApplyBeam step, this setting does not make sense (but it does if the applybeam is part of predict, ddecal, gaincal, h5parmpredict, etc.). Generally, ``false`` is the right setting for this option. The default has changed to false in Nov 2018 `.`


.. _ApplyBeam.json#/inputs/step_name:

step_name
+++++++++

:type: ``string``

:default: ``applybeam``

:doc: unique name for the step `.`


.. _ApplyBeam.json#/inputs/updateweights:

updateweights
+++++++++++++

:type: ``boolean?``

:doc: Update the weights column, in a way consistent with the weights being inverse proportional to the autocorrelations (e.g. if 'autoweights' was used before) `.`


.. _ApplyBeam.json#/inputs/usechannelfreq:

usechannelfreq
++++++++++++++

:type: ``boolean?``

:default: ``True``

:doc: Compute the beam for each channel of the measurement set separately. This is useful for merged / concatenated measurement sets. For raw LOFAR data you should set it to ``false``, so that the beam will be formed as in the station hardware. Also, setting it to false is faster `.`