Casa6

• 

  Josh Marvil 4 days ago 6cf1ede958f01d0f3ed691911c29f0f5c2778eb7

  CAS-14874: derive briggsbwtaper fracBW from output cube

  • casatools/src/code/synthesis/ImagerObjects/SynthesisImagerVi2.cc (version 6cf1ede958f01d0f3ed691911c29f0f5c2778eb7) (diffs)
• 

  Josh Marvil 4 days ago d27fff1705e9d9b7af956665dcd0a23eb5d5e692

  CAS-14874: Make the Briggs cube weight density partition-invariant
  The per-channel imaging-weight density that briggsbwtaper/perchanweightdensity
  relies on is local to a channel: f2/d2[chan] are derived from that channel's
  gridded-weight plane only.  The quality of that plane nevertheless depended on
  *how many* channels this process happens to own, which made the weights - and
  therefore the PSF and flux scale - vary with the number of MPI processes:
  
  - estimateSwingChanPad() scaled its extra margin by the sub-cube channel count
    (extrapad = max(min(4, imNChan/10), 1)), so different sub-cubes got different
    guard margins.
  - estimateSwingChanPad() returned 0 for a homogeneous frame, i.e. no guard
    margin at all, even though the FT machine still interpolates visibility
    frequencies across neighbouring channel planes (edge truncation).
  - init() padded its density template by a fixed +4/+2 channels while
    fillImgWeightCol() used the data-derived swingpad, so the two passes could
    disagree on the geometry.
  - the default pad for the multi-field/multi-MS path was a bare 4.
  - the scratch weight table key was (nrows, freqbeg, freqend, rmode, robust,
    interp).  All ranks chdir to the same working directory, so ranks that
    produced the same key shared one IMAGING_WEIGHT_* table and raced on it
    (reuse-if-nonzero-row-count, else delete-and-recreate) while it was being
    filled and consumed through readWeightColumn().  The reuse checks in
    initImgWeightCol()/init() keyed on (nx, ny, first-channel frequency) only and
    could hand back a density built for a different sub-cube.
  - initializeFTMachine() rebuilt grids_p[index] from the new template on every
    call but sized f2_p/d2_p only on the first one, so a template that grew
    between calls left f2_p/d2_p shorter than the grid and made
    f2_p[index][pos(3)] read past the end of the vector.
  
  Fix, in this file only:
  
  - introduce minWgtDensityEdgePad = 4 (channels per edge) and use it everywhere
    the guard margin was previously derived from the local channel count;
  - return that floor instead of 0 when no frame swing is present;
  - align init()'s template padding and the multi-field default with it;
  - fold a spectral geometry token (nchan, channel reference pixel, first
    channel frequency) into the scratch table name, and require it in the
    initImgWeightCol() early return;
  - pin the init() reuse on the cached density extent (f2_p) matching the
    requested template channel count;
  - grow f2_p/d2_p (and zero them) in initializeFTMachine() when the template
    channel count exceeds the cached allocation, and bounds-check the channel
    index in both weightUniform() and getWeightUniform() so a stale allocation
    yields a zero weight instead of an out-of-bounds read.
  
  Because the density itself is per-channel-local, making the geometry
  rule partition-independent is enough to make parallel results agree with serial
  for a single input MS, and it removes the shared-table race entirely.  No
  interface, XML or Python change.
  
  Known remaining partition dependence (follow-up, needs the full-cube channel
  count to be plumbed into this file):
  
  - the inOneGo threshold in initImgWeightCol() still compares the swing pad
    against templateimage.shape()[3]/10, so multi-MS inputs can still choose a
    different processing path on a small sub-cube than serial does;
  - estimateSwingChanPad() still derives freqbeg/freqend (and so the swing
    estimate itself) from the local template rather than the full selection, so
    the swing term remains weakly dependent on the sub-cube range;
  - the scratch table stays in the shared working directory (a per-rank workdir
    is not knowable here), so cross-run reuse of a same-tag, same-row-count table
    remains possible, as before.
  
  Validation: two standalone harnesses compile the changed logic against stubbed
  casacore types - the first asserts geometry-key/cache-key behaviour (reuse only
  for identical token/extent/nx/ny/freq) and the pad floor values, the second a
  faithful Block<Vector<Float>> model asserts the grow-then-zero path and that a
  stale shorter f2_p is rejected rather than read out of bounds.  Brace/paren
  balance preserved (110->112 braces, 794->849 parens, both sides equal).  A full
  casatools build plus serial-vs-parallel tests are still required before merge.

  • casatools/src/code/synthesis/TransformMachines2/BriggsCubeWeightor.cc (version d27fff1705e9d9b7af956665dcd0a23eb5d5e692) (diffs)