Unexplained discrepancy in Earth-Moon distance test value

[neilberkman]

While implementing Earth-Moon distance calculations in a port of skyfield to another language, I noticed a discrepancy in what appears to be a test reference value.

Simple km-to-AU conversion of the typical Earth-Moon distance:

>>> print(f"{384400 / 149597870.700:.15f}")  # typical Earth-Moon distance in km / AU in km  
0.002569555289800

However, I can't find the origin of the value 0.002568451424025 that appears in some tests. This value differs from the raw conversion by about 0.043% (ratio ≈ 1.0004298).

I've searched through the codebase but haven't found this exact value hardcoded. If this value includes corrections beyond the raw km-to-AU conversion, it would be helpful to document what those corrections are. If not, the test value may need to be updated to match the actual conversion.

[brandon-rhodes]

However, I can't find the origin of the value 0.002568451424025 that appears in some tests.

I don't see the value 0.002568451424025 in any of Skyfield's code — could you clarify where you have run across that value?

[brandon-rhodes]

As it's now been two months, I'm going to close this issue, as I don't see that number anywhere in Skyfield's code. But I'll be happy to answer further if you'd like to comment on the closed issue with a specific line number where I can find it.

[neilberkman]

Hey sorry for being non-responsive. I don't recall the issue at this point and I should have closed this out myself. Thanks for taking care of this.