Python wrapper for DPStep

DP3 offers advanced functionality that is field-tested in LOFAR production pipelines and, in some cases, not available otherwise. Making this functionality available to astronomers using SKA-LOW will be essential. Even commissioning scientists will benefit from a number of basic functions (flagging, DI calibration, averaging) provided by DP3.

• We have refactored all DPSteps such that they do no longer need an InputStep. This ensures that the DPSteps do not refer back to the original MeasurementSet to obtain the required meta-data. Instead, the DPSteps now fully rely on properly updated DPBuffer (data) and DPInfo (meta-data) objects, i.e., they retrieve data and meta-data from memory.
• We have created Python wrappers for the StepFactory (factory function to create a DPStep), DPBuffer and DPInfo, so that DPSteps can be created and called from Python.
• To illustrate the use of DPSteps from Python, we have developed a workflow script that performs direction-independent calibration from RASCIL.
• We have created a binary wheel to make the Python wrappers pip installable.
• DP3 provides a lot of functionality. Not all possible DPSteps and configurations of DPSteps are supported yet. However, as we have realised the underlying infrastructure, it should be straightforward to add additional functionality in the Python interface as we go along / as need arises.