Faster time coordinate categorization

[bouweandela]

🚀 Pull Request

Description

Make time coordinate categorization faster by converting the time points to datetimes in a single call.

Example script:

In [1]: import numpy as np

In [2]: import cf_units

In [3]: from iris.cube import Cube

In [4]: from iris.coords import DimCoord

In [5]: from iris.coord_categorisation import add_year

In [6]: coord = DimCoord(np.arange(10000), standard_name='time', units=cf_units.Unit("days since epoch", "standard"))

In [7]: cube = Cube(np.empty(10000), dim_coords_and_dims=[(coord, 0)])

In [8]: %timeit add_year(cube.copy(), 'time')
9.99 ms ± 24.1 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)

while with the current main branch, the timing result is

345 ms ± 2.07 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

---

Consult Iris pull request check list

---

Add any of the below labels to trigger actions on this PR:

• benchmark_this Request that this pull request be benchmarked to check if it introduces performance shifts

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 89.78%. Comparing base (9a23aa0) to head (16b20b3).
Report is 1 commits behind head on main.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #5999   +/-   ##
=======================================
  Coverage   89.77%   89.78%           
=======================================
  Files          88       88           
  Lines       23037    23056   +19     
  Branches     5038     5027   -11     
=======================================
+ Hits        20681    20700   +19     
  Misses       1624     1624           
  Partials      732      732           

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[bouweandela]

Would it make sense to add the example code above as a benchmark?

[trexfeathers]

Excellent idea @bouweandela, we just need to come to an agreement on how best to identify the cases where we're dealing with dates.

[trexfeathers]

OK, final couple of comments and I think we're ready to go 🚀