Tiny differences in enu2geodetic computation results between versions 2.8.0 and 3.1.0

[swiss-knight]

Hi,

I notice tiny differences in results of enu2geodetic computations between version 2.8.0 and 3.1.0 using the same input data:

v2.8.0

import pymap3d

pymap3d.enu2geodetic(48, 8, 600, 47.5, 8.3, 650, None, True)

Out [1]: (47.50007193912047, 8.300636957859766, 1250.0001852742141)

v3.1.0

import pymap3d
from pymap3d import Ellipsoid

ell = Ellipsoid.from_name("wgs84")

pymap3d.enu2geodetic(48, 8, 600, 47.5, 8.3, 650, None, True)

Out [1]: (47.50007193912046, 8.300636957859767, 1250.0001852742714)

Okay, it's only the last digit for lat and lon but the 3 last digits changed for the elevation.

Therefore, I'm currently wondering if the internal computations of the v 3.1.0 are more accurate or did anything else changed?

What are the exact WGS84 ellipsoid parameters used when setting ell=None (assuming the WGS84 ellipsoid is used in that case) in v.2.8.0 compared to its explicit definition using Ellipsoid.from_name("wgs84") v.3.1.0?

Thanks 🙏!