Resolution of Issue #71

This pull request fixes Issue #71, where aggte() fails after reloading a csdid result from a pickle file. The attached notebook provides a minimal replication of the failure and demonstrates the corrected behavior.

aggte_rowid_jupytext.ipynb

Root cause:

After reloading a saved csdid object, the internal data frame used in aggte_fnc/compute_aggte.py may contain a non-numeric column named rowid. During aggregation, the code groups the data and applies mean() across remaining columns, causing pandas to raise:

TypeError: agg function failed [how->mean, dtype->object]

Fix:

Before the aggregation step, the fix removes the rowid column if present:

if 'rowid' in data.columns:
data = data.drop(columns=['rowid'])

This restores correct execution of aggte() for both freshly computed and reloaded csdid objects, without affecting existing results.