Add Parquet export support for Clustergram widget

[cornhundred]

Summary

• add export_viz_parquet to Matrix for generating Parquet tables
• extend Clustergram widget with Parquet traitlets and support for parquet_data
• add helper for building network objects from Parquet on the frontend
• load Parquet data in widget render function when available

Testing

• npm run build (fails: cannot find package 'esbuild')
• npm run test (fails: jest not found)
• pytest (fails: unrecognized arguments --cov)

---

https://chatgpt.com/codex/tasks/task_b_6867f3185c98832a87f37669d6a56033

[Copilot]

Pull Request Overview

Adds full Parquet export support for the Clustergram widget by serializing cluster data server-side and deserializing it client-side.

• Introduce export_viz_parquet on Matrix to emit Parquet-encoded byte blobs for all viz tables
• Extend Clustergram widget to accept a parquet_data or fallback matrix and expose new Bytes traitlets
• Add a front-end helper (networkFromParquet) and update render_matrix_new to rebuild the network from Parquet bytes

Reviewed Changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated no comments.

Show a summary per file

File	Description
src/celldega/clust/matrix.py	Add export_viz_parquet method to serialize clustering tables to Parquet
src/celldega/viz/widget.py	Extend Clustergram constructor: handle parquet_data, add network_meta and dynamic Byte traits
js/read_parquet/network_from_parquet.js	New helper to convert Arrow tables back into the network object
js/widget.js	Update render_matrix_new to detect Parquet bytes and use networkFromParquet
notebooks/Single_Category_Matrix.ipynb	Demo notebook showing Parquet-based widget usage

Comments suppressed due to low confidence (4)

src/celldega/clust/matrix.py:594

• Add unit tests for export_viz_parquet covering both clustered and non-clustered matrices and verify the shapes and contents of the returned byte arrays.

    def export_viz_parquet(self) -> dict[str, bytes]:

src/celldega/clust/matrix.py:595

• [nitpick] Expand this docstring to list each key in the returned dict (mat, row_nodes, col_nodes, row_linkage, col_linkage, meta) and describe their formats and meaning.

        """Export visualization using Parquet encoded tables."""

src/celldega/viz/widget.py:125

• [nitpick] Add or update the constructor docstring to clarify the precedence and expected behavior when network, parquet_data, and matrix are all provided.

        pq_data = kwargs.pop("parquet_data", None)

js/read_parquet/network_from_parquet.js:1

• Add automated tests for networkFromParquet to confirm correct reconstruction of the network object from Arrow tables, including edge cases like empty or malformed byte inputs.

import { arrayBufferToArrowTable } from './arrayBufferToArrowTable';