GainCal
Calibrate gains using StefCal.
inputs
type
- type:
string- doc:
Case-insensitive step type; must be ‘gaincal’ or ‘calibrate’ .
applybeamtomodelcolumn
- type:
bool- default:
False- doc:
Apply the beam model (at the phase center) to the visibilities in the model column. If this option is true, all options from ApplyBeam are valid as well (except .invert, since the model data will always be corrupted for the beam) .
applycal.*
- doc:
ApplyCal sub-step, same as in Predict step .
applysolution
- type:
bool- default:
False- doc:
Apply the calibration solution to the visibilities. Note that you should always also inspect the parmdb afterwards to check that the solutions look reasonable .
baseline
- type:
string- doc:
Baseline selection filter for calibration. See Description of baseline selection parameters. New in version 2.20 .
beammode
- type:
enum- doc:
Same as in ApplyBeam step .
- symbols:
array_factor,element,default
blrange
- type:
vector- doc:
Vector of baseline lengths to use for calibration. See Description of baseline selection parameters. New in version 2.20 .
caltype
- type:
string- default:
diagonal- doc:
The type of calibration that needs to be performed. Options are scalar, scalarphase, scalaramplitude, diagonal, diagonalphase, diagonalamplitude, fulljones, ‘tec’ and ‘tecandphase’ .
debuglevel
- type:
int- default:
0- doc:
Debugging. If debuglevel==1, then a file debug.h5 is created containing all iterands. This file will be very large; you can use it to check the convergence speed etc .
detectstalling
- type:
bool- default:
True- doc:
Detect if the iteration does not converge anymore and then stop iterating even if maxiter is not reached .
maxiter
- type:
int- default:
50- doc:
Maximum number of iterations of stefcal .
minblperant
- type:
int- default:
4- doc:
If an antenna has less than minblperant unflagged data points for a given solution slot, it is not used for calibration .
modelcolumn
- type:
string- default:
MODEL_DATA- doc:
Name of the model data column in the input measurement set .
nchan
- type:
int- default:
0- doc:
Number of channels on which a solution is assumed to be constant (same as CellSize.Freq in BBS). 0 means all channels. When caltype = ‘tec’ or ‘tecandphase’, the default is 1, meaning that a TEC will be fitted through a phase for each channel .
onebeamperpatch
- type:
string?- doc:
Same as in ApplyBeam step .
parmdb
- type:
string- doc:
Path of parmdb in which the computed parameters are to be stored. If the parmdb already exists, it will be overwritten. Note: You cannot use this parmdb in an applycal step in the same run of DP3. To apply the solutions of the gaincal directly, use ‘gaincal.applysolution’ (see below). New in LOFAR 3.1: if the parmdb name ends in
.h5, an H5Parm will be written .
propagatesolutions
- type:
bool- default:
True- doc:
Use solutions of one time interval as a starting value for the next time interval .
reusemodel
- type:
string- doc:
Reuse model data from previous steps. The string should contain the model data name, as set by a previous step, e.g., ddecal.center .
solint
- type:
int- default:
1- doc:
Number of time slots on which a solution is assumed to be constant (same as CellSize.Time in BBS). 0 means all time slots. Note that for larger settings of solint, and specially for solint = 0, the memory usage of gaincal will be large (all visibilities for a solint should fit in memory) .
sourcedb
- type:
string- doc:
Same as in Predict step .
sources
- type:
string?- default:
[]- doc:
Same as in Predict step .
timeslotsperparmupdate
- type:
int- default:
500- doc:
Number of solution intervals after which the parmdb is updated .
tolerance
- type:
float- default:
1e-05- doc:
Tolerance to which the model should match the data .
usebeammodel
- type:
bool- default:
False- doc:
Same as in Predict step .
usechannelfreq
- type:
boolean?- doc:
Same as in ApplyBeam step .
usemodelcolumn
- type:
bool- default:
False- doc:
Use model column. The modelcolumn setting contains the model column name .
uvlambdamin
- type:
double- default:
0- doc:
Ignore baselines / channels with UV < uvlambdamin wavelengths. Note: also all other variants of uv flagging described in UVWFlagger (uvmmin, uvmrange, uvlambdarange, etc) are supported (New in 3.1) .