FgcmCalibrateTractTableTask¶
FgcmCalibrateTractTableTask
will run the full fgcmcal
processing cycle on a single tract with multi-band coverage.
Note that the results will not be as robust as a full global calibration because of the limited ability to track instrumental changes from observations in a single tract.
Running FgcmCalibrateTractTableTask
requires a look-up table generated by FgcmMakeLutTask.
FgcmCalibrateTractTableTask
uses FgcmBuildStarsTableTask to extract stars from sourceTable_visit
parquet tables.
Processing summary¶
FgcmCalibrateTractTableTask
runs the full fgcmcal
processing, from building star lists to fitting to the output of final products.
Specifically, it will:
Build the star lists: FgcmBuildStarsTableTask
Run the fitter until convergence: FgcmFitCycleTask
Output the final data products: FgcmOutputProductsTask
Python API summary¶
from lsst.fgcmcal.fgcmCalibrateTractTable import FgcmCalibrateTractTableTask
-
class
FgcmCalibrateTractTableTask
(initInputs=None, **kwargs) Calibrate a single tract using fgcmcal, using sourceTable_visit (parquet) input catalogs
...
- attributeconfig
Access configuration fields and retargetable subtasks.
-
method
run
(handleDict, tract, buildStarsRefObjLoader=None, returnCatalogs=True) Run the calibrations for a single tract with fgcm
...
See also
See the FgcmCalibrateTractTableTask
API reference for complete details.
Retargetable subtasks¶
fgcmBuildStars¶
Task to load and match stars for fgcm
fgcmOutputProducts¶
- Default
- Field type
Task to output fgcm products
Configuration fields¶
connections¶
- Data type
lsst.pipe.base.config.FgcmCalibrateTractTableConfigConnections
- Field type
Configurations describing the connections of the PipelineTask to datatypes
convergenceTolerance¶
Tolerance on repeatability convergence (per band)
doDebuggingPlots¶
Make plots for debugging purposes?
fgcmFitCycle¶
- Data type
- Field type
Config to run a single fgcm fit cycle
maxFitCycles¶
Maximum number of fit cycles
saveLogOutput¶
Flag to enable/disable saving of log output for a task, enabled by default.