Fix #466 by correcting from_satrec() epoch formula

skyfielders · Oct 31, 2020 · 4bcdecb · 4bcdecb
1 parent 9a718e7
commit 4bcdecb
Show file tree

Hide file tree

Showing 3 changed files with 10 additions and 1 deletion.
diff --git a/CHANGELOG.rst b/CHANGELOG.rst
@@ -2,6 +2,14 @@
 Changelog
 =========
 
+1.32 — Future
+-------------
+
+* Fix: the ``.epoch`` attribute of Earth satellite objects that were
+  built using :meth:`~skyfield.sgp4lib.EarthSatellite.from_satrec()`
+  was, alas, a half-day off.
+  `#466 <https://github.com/skyfielders/python-skyfield/issues/466>`_
+
 1.31 — 2020 October 24
 ----------------------
 

diff --git a/skyfield/documentation/earth-satellites.rst b/skyfield/documentation/earth-satellites.rst
@@ -740,7 +740,7 @@ call this special constructor:
 .. testoutput::
 
     Satellite number: 5
-    Epoch: A.D. 2000-Jun-27 06:50:24.0000 UT
+    Epoch: A.D. 2000-Jun-27 18:50:24.0000 UT
 
 The result should be a satellite object that behaves
 exactly as though it had been loaded from TLE lines.
diff --git a/skyfield/sgp4lib.py b/skyfield/sgp4lib.py
@@ -131,6 +131,7 @@ def from_satrec(cls, satrec, ts):
         # be unified to always use epochyr and epochdays.
         whole, fraction = divmod(satrec.jdsatepoch, 1.0)
         year, month, day = compute_calendar_date(whole)
+        day += 0.5  # convert integer Julian day into Julian date float
         self.epoch = ts.utc(year, month, day + fraction + satrec.jdsatepochF)
 
         self._setup(satrec)