Bug: _ekadhipatya_sodhana only reduces each planet's own sign-pair and skips Sun & Moon

[Askthesages]

Version: PyJHora 4.8.7
File: src/jhora/horoscope/chart/ashtakavarga.py → _ekadhipatya_sodhana

Summary

Ekadhipatya Sodhana should reduce every graha's bhinnashtakavarga across all the co-lordship (ekadhipatya) sign-pairs. The current implementation reduces only the diagonal: each planet's chart is reduced solely at that same planet's own owned pair, and the Sun and Moon charts are never reduced at all.

Cause

One loop index is used for two different things:

for p in range(const.MOON_ID+1, const.SATURN_ID+1):   # p = Mars..Saturn
    r1, r2 = rasi_owners[p]                            # the pair OWNED by planet p
    ...
    bav[p][r1] ...                                     # but bav[p] = planet p's OWN BAV row

So bav[Mars] is reduced only at Aries/Scorpio, bav[Mercury] only at Gemini/Virgo, etc., and bav[Sun] / bav[Moon] are skipped entirely.

Effect

Sodhya Pinda comes out wrong for most charts. Example (1978-11-22, 11:40, Rio): after trikona reduction, the Sun BAV keeps non-zero bindus in Gemini and Virgo that the classical all-pairs reduction removes, so Sun's Sodhya Pinda differs from the correct value.

Expected

Reduce each graha's BAV (Sun..Saturn, including Sun and Moon) across all five dual-owner pairs:

for p in const.SUN_TO_SATURN:                            # every graha's BAV, incl. Sun & Moon
    for r1, r2 in [rasi_owners[i] for i in range(2, 7)]:  # all 5 dual-owner pairs
        # apply the same per-pair reduction rule here

Happy to send a PR if useful.

[naturalstupid]

The implementation of def _ekadhipatya_sodhana was based on PVR's book Section 12.7.2 Page.158

[naturalstupid]

Sorry. It took me sometime to understand your changes. I had been observing my PyJHora results were not matching PVR's Book Exercise values. I made your changes now. (I dont know how to use pull requests - so I added your changes manually in next version with the comment to the issue number your github-id: Askthesages. Still I was not getting the book results. I had been assuming something wrong with the print.
After searches across various other sites - I noticed two things were either missing or inaccurate.

1. Rule-3 in above book chapter - does not include the case where both the values are equal. Other sources indicate in such case - empty rasi is reduced to zero.
2. rasimana multipliers for sodhaya pinda given in PVR's book is [7, 10, 8, 4, 10, 6, 7, 8, 9, 5, 11, 12] - whereas many other sources say rasimana multiplier for Virgo should be 5 and not 6. that is rasimana multiplers = [7, 10, 8, 4, 10, 5, 7, 8, 9, 5, 11, 12]

With the above changes now the book results are matching. With respect to rasimana multiplier since I was not sure - which source is correct - I am adding the rasimana_option argument defaulted to other sources and users can change configuration in config UI dialog.
here is the revised function

def _ekadhipatya_sodhana(binna_ashtaka_varga_after_trikona, chart_1d):
    """
        V4.8.7 - Changes per Issue #52 from "Askthesages"

        Note:
        In Ekaadhipatya Sodhana Rule (3), the printed text commonly states:
          - if the empty rasi has a lower value -> set it to 0
          - if the empty rasi has a higher value -> set it equal to the occupied rasi
        but it does not explicitly mention the equal-value case.

        Following several secondary Ekadhipatya references, when one rasi is occupied
        and the other is empty, and both have equal values, the empty rasi is also
        reduced to zero. This reproduces the published Chart 7 SoAV/Sodhya-Pinda
        values when used with the alternate Rasimana multipliers. 
        See e.g. Horosoft / AstroVastuTips style explanations of Ekadhipatya Shodhana.
    """
    bav = [row[:] for row in binna_ashtaka_varga_after_trikona]

    rasi_owners = const.ashtakavarga_rasi_owners
    dual_pairs = [x for x in rasi_owners if isinstance(x, tuple)]  # the 5 co-owned (r1,r2) pairs

    for p, (r1, r2) in [(_p, _pr) for _p in const.SUN_TO_SATURN for _pr in dual_pairs]:
        r1_occupied = not (chart_1d[r1].strip() == '')
        r2_occupied = not (chart_1d[r2].strip() == '')

        # Rule 1: if either rasi has zero, no reduction.
        # Rule 2: if both rasis are occupied, no reduction.
        if (bav[p][r1] == 0 or bav[p][r2] == 0) or (r1_occupied and r2_occupied):
            continue

        elif (not r1_occupied) and (not r2_occupied):  # Rule 4 - both rasis empty
            # Rule 4(b): if different, replace the higher value with the lower value.
            if bav[p][r1] != bav[p][r2]:
                min_value = min(bav[p][r1], bav[p][r2])
                bav[p][r1] = min_value
                bav[p][r2] = min_value
            # Rule 4(a): if equal, replace both with zero.
            else:
                bav[p][r1] = 0
                bav[p][r2] = 0

        else:
            # Rule 3: one rasi occupied, the other empty.
            # The PVR Book lists only "lower" and "higher" cases for the empty rasi
            # and omits the equal-value case.
            #
            # Following secondary Ekadhipatya references, if the occupied and empty
            # rasis have equal values, the empty rasi is also reduced to zero.
            # In short:
            #   empty < occupied  -> empty = 0
            #   empty = occupied  -> empty = 0
            #   empty > occupied  -> empty = occupied

            if r1_occupied:  # r2 is empty
                if bav[p][r2] <= bav[p][r1]:
                    bav[p][r2] = 0
                else:
                    bav[p][r2] = bav[p][r1]

            else:  # r1 is empty
                if bav[p][r1] <= bav[p][r2]:
                    bav[p][r1] = 0
                else:
                    bav[p][r1] = bav[p][r2]

    return bav
```

Let me know you are ok with changes or any other comments on this topic.