explore

Agnostic dataset and experiment exploration.

Lightweight introspection for both pre-load discovery results (datakit.Dataset) and post-load materialized outputs (pandas.DataFrame). Output uses rich when available and falls back to plain indented text otherwise.

Quick start:

from mesofield.datakit import explore, Dataset

# Pre-load: inspect a discovered Dataset (or directory path)
explore("path/to/experiment")
explore(Dataset.from_directory(root))

# Post-load: inspect a materialized DataFrame or pickle / HDF5
explore("path/to/dataset.pkl")
explore(materialized_df)

# Programmatic access (no printing)
report = explore(dataset, print_output=False)