add a few benchmarks

[dougbrn]

Resolves #37. Adds a few more benchmarks. I opted to just test add_nested instead of adding parquet files for now, as in principle our package inherits the performance of parquet reading from pandas. It may still be good to add in the future (or now if others feel strongly) to check for any regressions, but it would be slightly redundant with whatever testing Pandas itself already does.

[github-actions]

Before [ab5fc4c]	After [c2b2977]	Ratio	Benchmark (Parameter)
8.24±0.2ms	8.54±0.08ms	1.04	benchmarks.NestedFrameQuery.time_run
32.0±2ms	32.7±3ms	1.02	benchmarks.AssignSingleDfToNestedSeries.time_run
5.12±0ms	5.16±0.02ms	1.01	benchmarks.NestedFrameReduce.time_run
85.8M	86M	1	benchmarks.NestedFrameAddNested.peakmem_run
8.14±0.06ms	8.12±0.06ms	1	benchmarks.NestedFrameAddNested.time_run
88.1M	88.4M	1	benchmarks.NestedFrameQuery.peakmem_run
88.1M	88.4M	1	benchmarks.NestedFrameReduce.peakmem_run
271M	272M	1	benchmarks.ReassignHalfOfNestedSeries.peakmem_run
257M	255M	0.99	benchmarks.AssignSingleDfToNestedSeries.peakmem_run
59.0±3ms	58.4±3ms	0.99	benchmarks.ReassignHalfOfNestedSeries.time_run

Click here to view all benchmarks.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 97.55%. Comparing base (ab5fc4c) to head (e71a2c7).

Additional details and impacted files

@@           Coverage Diff           @@
##             main      #58   +/-   ##
=======================================
  Coverage   97.55%   97.55%           
=======================================
  Files          16       16           
  Lines         818      818           
=======================================
  Hits          798      798           
  Misses         20       20           

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

[hombit]

Looks great