Built-in operations

There are a small number of built-in Operations provided. Some of them are quite general-purpose, which some others were developed for a specific radio-astronomy pipeline and might not be useful elsewhere, but could be seen as examples.

At present there is no step-by-step user guide for these operations, so you will need to refer to the linked API docs to learn how to use them. The links below are to the operation template classes, which contain the high-level descriptions of the operations; the corresponding operation classes document the slots they provide.

• FillTemplate: fill a buffer with a constant value.

• TransposeTemplate: transpose a 2D array.

• HReduceTemplate: apply a reduction (such as a sum) to each row of a 2D array.

• MaskedSumTemplate: compute a weighted sum of each column of a 2D array.

• Percentile5Template: compute percentiles of the data in each row of a 2D array.

RFI flagging

The katsdpsigproc.rfi.device module contains a number of classes for detecting RFI (Radio Frequency Interference) in radio-astronomy data. There are also CPU-only equivalents of each of these classes (mostly for testing rather than high performance) in katsdpsigproc.rfi.host. There is also a more sophisticated and optimized flagger in katsdpsigproc.rfi.twodflag (so-called because it looks for anomalies in both time and frequency).