Update read_parquet_file_to_pandas to use nested pandas I/O

[smcguire-cmu]

No description provided.

[codecov]

Codecov Report

Attention: Patch coverage is 80.00000% with 1 line in your changes missing coverage. Please review.

Project coverage is 92.32%. Comparing base (5d13465) to head (5813ef7).
Report is 4 commits behind head on main.

Files with missing lines	Patch %	Lines
src/hats/io/file_io/file_io.py	80.00%	1 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #499      +/-   ##
==========================================
+ Coverage   92.23%   92.32%   +0.08%     
==========================================
  Files          47       47              
  Lines        1984     1993       +9     
==========================================
+ Hits         1830     1840      +10     
+ Misses        154      153       -1     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[github-actions]

Before [dc1aff8]	After [d107289]	Ratio	Benchmark (Parameter)
13.0±0.3ms	13.6±0.6ms	1.05	benchmarks.Suite.time_inner_pixel_alignment
37.4±0.2ms	39.1±0.2ms	1.05	benchmarks.Suite.time_pixel_tree_creation
18.7±0.1ms	18.8±0.2ms	1.01	benchmarks.MetadataSuite.time_load_partition_info_order6
355±2ms	355±3ms	1	benchmarks.Suite.time_outer_pixel_alignment
1.05±0.01ms	1.05±0.01ms	1	benchmarks.time_test_cone_filter_multiple_order
74.5±1ms	73.6±0.5ms	0.99	benchmarks.MetadataSuite.time_load_partition_info_order7
74.8±0.8ms	73.8±0.4ms	0.99	benchmarks.MetadataSuite.time_load_partition_join_info
111±1ms	110±0.8ms	0.99	benchmarks.time_test_alignment_even_sky

Click here to view all benchmarks.

[delucchi-cmu]

See astronomy-commons/hats-cloudtests#71

[dougbrn]

Is serialization for parquet writing on the docket for this, or is the plan to do that in a later PR (or does it already work? If you're writing closer to the pyarrow side then I suppose there's a decent chance it does)

[smcguire-cmu]

Is serialization for parquet writing on the docket for this, or is the plan to do that in a later PR (or does it already work? If you're writing closer to the pyarrow side then I suppose there's a decent chance it does)

Yeah, in LSDB we just use df.to_parquet, so I think writing should work fine with that.