Changelog
1.1.0
Support MSv2 as input to pipeline.
Upgrade xradio to 1.1.3
1.0.0
Upgrade radler dependency to 1.0
Upgrade aoflagger dependency to 3.5
Align frequency dimension value in predict processing function
0.8.0
Migrate to ska-sdp-exec-piper V2 API
0.7.0
Add support for numpy 2+
Make use of radler deconvolver optional by default
Make flagging stage optional by default
Reduce memory usage while performing power law scaling in predict stage
0.6.4
Remove internal piper package, and depend on
ska-sdp-exec-piperrepositoryUpdate load_data stage to allow reading from multiple partitions of MSv4 data
Reduce memory usage in
read_modelstageAdd a script to benchmark the pipeline
0.6.3
Fix xarray coordinate mismatch in deconvolver step
Update minimum dask and xarray versions
Rename cli command to
ska-sdp-spectral-line-imaging
0.6.2
Add support for power law scaling in
read_modelstageAdd support for configurable spectral line imaging flaggin strategy
Output FITS images are now written to seperate file per polarisation
0.6.1
Add
flagging_stage.The
read_modelstage can accept a FITS spectral cube as a model image.The
read_modelandimagingstage parameters have changed. Please refer the "Stage Configs" section of the documentation.The
cont_substage reports the peak visibility in subtracted visibilities and the corresponding channelThe
cont_substage can optionally report extent of polynomial fit, based on input configurationUse
subtract_visibilityandconvert_polarizationfromska-sdp-func-pythonRemove
input_polarisation_framefromvis_stokes_conversionstageRename
output_polarisation_frametooutput_polarizations, this option will consume an array of string for converting polarization framesAdd dask worker plugin to configure logger
Introduce delayed logger, used to log values which need computations
Remove all export stages. The export happens from the respective stages where the data is generated.
Move configs related to exported image name and image format to imaging stage, with the addition of export flags for image, psf, model, and residual.
Move read processing set logic from piper framework to spectral line imaging pipeline
Delay reading of FITS model images till execution of graph
Add support for power law scaling of continuum model images
Add
benchmarksubcommand inpiperwhich can be used to benchmark pipeline usingdoolUpdate xradio version to 0.0.40
Add auto-complete script for bash and zsh
0.5.0
Update documentation
Add support for radler deconvolver
Add script to install config and run pipeline
0.4.0
Implement clean algorithm using ducc gridder and hogbom deconvolver
Allow pipeline to write spectral cube in FITS format
Add functionality to estimate cell size and image size for imaging
Add option
--setto sub-commandinstall-configto override configFix vulnerabilities in docker image
Performance improvements in the diagnostic stage
0.3.0
Add Singularity Image
Update Documentation
Update default pipeline configuration
0.2.0
Documentation updates
Enable dask bench marking
CSD3 example script
Move linting dependencies to dev section in pyproject
Streamline dev environment setup
0.1.0
Initial release of SKA SDP Spectral Line Imaging Pipeline