Difference in the definition of "First Day of the Month"?

[floydpink]

The difference in the definition of the first day of the month (mentioned in passing on p.14 - p.15 in this paper - see screenshot below for the excerpt) is causing the discrepancy (as compared to the Saka solar month and day) in the Kollavarsham dates of some months.

Consider this example:
kollavarsham -b 16:02:2014 yields this:

Output
------
Saka : 1935 02 Magha      (10)  [Krsnapaksa]
Saura: 1935 04 Kumbha     (10)  P-phalguni
ME   : 1189 04 Kumbham    (06)  Pooram

And this seems to match with the Deepika calendar for 2014

But this command's output does not match the day alone with the calendar:
kollavarsham -b 16:03:2014

Output
------
Saka : 1935 15 Phalguna   (11)  [Suklapaksa]
Saura: 1935 03 Mina       (11)  P-phalguni
ME   : 1189 03 Meenam     (07)  Pooram

@arunkutty Thoughts?

[arunkutty]

There are other dates also where this happens. In borderline cases the algorithm seems to be inaccurate. The modern calendar day has most of the times multiple nakshatras spread out in it. I think for a Kollavarsham conversion to be accurate the time of the day also plays an important role.

[floydpink]

True.

The multiple nakshatras on a single day scenario as well as multiple occurrence of the same nakshatra in a single Malayalam month has repercussions in when the pirannalu is considered.

[arunkutty]

I think this might hold true for all events and not only pirannal. We might need to sort it out for all practical applications of the api.