doc

CHANGELOG-unreleased.md

@@ -13,6 +13,7 @@ the released changes.
 - Moved design matrix normalization code from `pint.fitter` to the new `pint.utils.normalize_designmatrix()` function.
 - Made `Residuals` independent of `GLSFitter` (GLS chi2 is now computed using the new function `Residuals._calc_gls_chi2()`).
 - Optionally avoid certain computations in `Astrometry.get_d_delay_quantities()` method.
+- Removed an unnecessary for loop in `solar_system_shapiro_delay`
 ### Added
 - Added `WaveX` model as a `DelayComponent` with Fourier amplitudes as fitted parameters
 - `Parameter.as_latex` method for latex representation of a parameter.

src/pint/models/astrometry.py

@@ -63,14 +63,15 @@ def ssb_to_psb_xyz_ICRS(self, epoch=None):
         # TODO: would it be better for this to return a 6-vector (pos, vel)?
 
         # @abhisrkckl: The following operations are stupid slow for some reason.
-        # This is somewhat uunderstandable when the proper motion is non-zero, and
+        # This is somewhat understandable when the proper motion is non-zero, and
         # we'll have to do this for every epoch. But this is slow even when there is
         # no proper motion. The underlying operations should just be some trig functions
         # and simple 3x3 matrix operations; so I don't understand why this has to be so slow.
         # Anyway, I am not going down this rabbit hole right now. This rant is here to remind
-        # people that there is a performance bottleneck here. Same problem is there with a similar
-        # function in AstrometryEcliptic.
-        # As a stopgap measure, I am moving this computation to a cached pure function below.
+        # people that there is a performance bottleneck here.
+        #
+        # As a stopgap measure, I am overriding this function in the subclasses without
+        # using `astropy.coordinates`.
 
         return self.coords_as_ICRS(epoch=epoch).cartesian.xyz.transpose()