improving duckdb + autopandas performance

[edublancas]

Describe your changes

• improved performance with using duckdb and autopandas (via DuckDB's native .df())
• added notebook with DuckDB benchmarks
• moved ResultSet constructor to the top
• refactored ResultSet, no longer a subclass of list (using composition instead of inheritance)
• added missing ResultSet tests

Issue number

Closes #451

Checklist before requesting a review

• Performed a self-review of my code
• Formatted my code with pkgmt format
• Added tests (when necessary).
• Added docstring documentation and update the changelog (when needed)

---

📚 Documentation preview 📚: https://jupysql--469.org.readthedocs.build/en/469/

[idomic]

Please fix

[yafimvo]

@edublancas Nice improvement! ~6.5 sec vs ~260 ms

Should we see the benchmark in the docs?

[idomic]

Should we see the benchmark in the docs?

It's intentionally not part of the docs, as this is how it should operate regardless (let's say without JupySQL).