Issue with Julian Date Conversion of BC date?

I was trying to find the time and angular separation for Venus and Jupiter for the famous ‘merging’ conjunction on June 17, 2 BC, and when I try to find it using skyfield, the date appeared off by 2 days using the de422.bsp ephemeris.

But after some investigation, I think this could be an issue converting the Julian Date. I say this because the conjunction was found at JD 1720860.3694220667 and I suspect this JD is sound. One online calculator reported this as June 17, 2 BC (the date I was expecting), but executing the following code:

import scipy.optimize
from skyfield.api import load, pi, tau, Topos

ts = load.timescale()
jd_conjunction = 1720860.3694220667
conjunc_time= ts.tt(jd=jd_conjunction)
conjunc_time_str = conjunc_time.utc_jpl()

print(f'Converted {jd_conjunction} to {conjunc_time_str}')

returns Converted 1720860.3694220667 to B.C. 0002-Jun-15 20:51:15.8826 UT

I will note the June 15 date does match the AAVSO converter (assuming year -1 = 2 BC), so it may be a common algorithm error on someone’s part (either skyfield and AAVSO share an error or other converters share an error). Most sky simulation software shows the conjunction on June 17, 2 BC.

[JuanCab]

Went ahead and checked, that was the problem, consider this issue closed.

[JuanCab]

Oh, of course… OK, I’ll take a look at it tomorrow, but now it makes more sense.