@onekiloparsec onekiloparsec released this Jul 29, 2017 · 169 commits to master since this release

Assets 2

The problem was a confusion between two types of ecliptic coordinates (one geocentric, one heliocentric). And the equatorial coordinates needed for the computation of rise transit set times were based on the wrong ones (heliocentric).

All tests pass. But for some of them, the accuracy isn't satisfactory (> a few minutes, up to some hours). With the large spread of sources on the Internet, often providing little clue on how things are computed, make it very hard here to have a consistent set of tests which are accurate compared to all these sources at the same time.

Nevertheless, we keep thriving!

For the archive, let state that the sources of data for tests are: The AA book (by Jean Meeus), AA+ own C++ tests, USNO, Sky Safari, Xephem.