Fixes for accessor mapping methods

[hombit]

This PR addresses issue #87 and removes most of the in-place modification methods from the NestedAccessor. The problem with field modification was that pd.Series.dtype would not update after changing the underlying NestedAccessor.dtype. The only modification we can still perform through the accessor interface is changing the field (subcolumn) data without altering its dtype.

Note: When I mention "in-place" here, I don't mean that we do not copy data — we do copy the arrays we modify. However, this process is transparent from the user's and pd.Series's point of view. Therefore, we are effectively changing pd.Series in-place, while the underlying memory blobs remain immutable.

It also fixes #62, because we wouldn't have both methods anymore.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Project coverage is 98.76%. Comparing base (41cce93) to head (d2c4ebd).
Report is 72 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main      #91      +/-   ##
==========================================
+ Coverage   98.71%   98.76%   +0.04%     
==========================================
  Files          15       15              
  Lines         859      888      +29     
==========================================
+ Hits          848      877      +29     
  Misses         11       11              

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

[github-actions]

Before [41cce93]	After [fbacf97]	Ratio	Benchmark (Parameter)
9.37±0.1ms	9.61±0.2ms	1.03	benchmarks.NestedFrameAddNested.time_run
276M	280M	1.02	benchmarks.ReassignHalfOfNestedSeries.peakmem_run
61.6±3ms	62.1±2ms	1.01	benchmarks.ReassignHalfOfNestedSeries.time_run
86.2M	86.2M	1	benchmarks.NestedFrameAddNested.peakmem_run
89.4M	89.4M	1	benchmarks.NestedFrameQuery.peakmem_run
89.4M	89.4M	1	benchmarks.NestedFrameReduce.peakmem_run
6.56±0.2ms	6.46±0.2ms	0.99	benchmarks.NestedFrameQuery.time_run
5.59±0.1ms	5.53±0.3ms	0.99	benchmarks.NestedFrameReduce.time_run
261M	256M	0.98	benchmarks.AssignSingleDfToNestedSeries.peakmem_run
33.8±1ms	31.1±2ms	0.92	benchmarks.AssignSingleDfToNestedSeries.time_run

Click here to view all benchmarks.

[dougbrn]

Looks great!