orca.transform.spectrum_v3
==========================

.. py:module:: orca.transform.spectrum_v3

.. autoapi-nested-parse::

   Dynamic-spectrum helpers (v3)
   * 100% read-only MeasurementSet access
   * Optionally respects MS FLAG column
   * bcal optional; if present, used at native resolution (NO rebin)
   * NO spectral averaging: 192 chans/subband → 16 SPWs → 3072 total chans



Attributes
----------

.. autoapisummary::

   orca.transform.spectrum_v3.logger
   orca.transform.spectrum_v3.N_CHAN_OUT
   orca.transform.spectrum_v3.REDIS_URL
   orca.transform.spectrum_v3.REDIS_EXPIRE_S
   orca.transform.spectrum_v3.REDIS_KEY_PREFIX
   orca.transform.spectrum_v3.ROW_NUMS


Functions
---------

.. autoapisummary::

   orca.transform.spectrum_v3.dynspec_map_v3
   orca.transform.spectrum_v3.dynspec_reduce_v3


Module Contents
---------------

.. py:data:: logger

.. py:data:: N_CHAN_OUT
   :value: 192


.. py:data:: REDIS_URL

.. py:data:: REDIS_EXPIRE_S
   :value: 36000


.. py:data:: REDIS_KEY_PREFIX
   :value: 'spec-v3-'


.. py:data:: ROW_NUMS
   :value: [('LWA-128&LWA-160', 54282), ('LWA-048&LWA-051', 33335), ('LWA-018&LWA-237', 30360),...


.. py:function:: dynspec_map_v3(subband_no: int, scan_no: int, ms: str, bcal: str | None = None, use_ms_flags: bool = True) -> List[_SnapshotSpectrumV3]

   Map step on a SINGLE MS:
     • Reads DATA & FLAG (cross-correlations only)
     • Optionally applies bcal at native resolution (192 channels) — NO rebin
     • Stores incoherent-sum and selected-baseline spectra in Redis
   Returns JSONable _SnapshotSpectrumV3 list (for the reducer).


.. py:function:: dynspec_reduce_v3(spectra: Iterable[List[_SnapshotSpectrumV3]], start_ts: datetime.datetime, out_dir: str) -> None

   Reduce step:
     • Gathers Redis blobs from dynspec_map_v3 outputs
     • Builds 4×(time×freq) cubes per type, freq = 192×16 = 3072
     • Writes FITS into {out_dir}/{type}/{DATE}-{corr}.fits