ENH: Mapping: Small performance fix

[richardotis]

This PR moves FastFunctionFactory.getfunc() out of a tight loop inside of PhaseRecord.prop_2d(). Despite hitting the cache, it seems that getfunc() takes appreciable time and, for large numbers of inputs (num_imps), it can become as much as 40% of the runtime of that function. For large values of pdens (such as the value 500, used by default inside of mapping), this significantly impacts the calculate function.

On my laptop, this change shaves a few seconds off the mapping tests.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 91.75%. Comparing base (c24d881) to head (8c198a4).

Additional details and impacted files

@@           Coverage Diff            @@
##           develop     #552   +/-   ##
========================================
  Coverage    91.74%   91.75%           
========================================
  Files           77       77           
  Lines        11981    11982    +1     
========================================
+ Hits         10992    10994    +2     
+ Misses         989      988    -1     

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