Just say the magic word.

It pays to be polite.

One function, 40+ adapters, zero required dependencies. Pass any object—DataFrames, tensors, plots, sklearn pipelines—and get a clean summary plus structured metadata. Useful for debugging, fast exploration, and those moments where you forgot what your object even is.

Install

Zero required dependencies. Opt into extras when you're ready.

pip install pretty-little-summary

Ask nicely.

One function. Pass it anything. It figures out the rest.

import pretty_little_summary as pls

# The object screaming errors at you
messy_data = load_nightmare_csv()

# print(messy_data)  ← Rude. Useless.
# messy_data.info()  ← Too much drama.

# Try asking nicely:
result = pls.describe(messy_data)
print(result.content)

What you get back

A Summary with human-readable content and structured metadata.

Structured metadata

result.meta
# {"object_type": "pandas.DataFrame",
#  "shape": [3, 3],
#  "columns": ["product", "price", "quantity"]}

Example gallery

Browse real inputs and the exact summaries they produce. Search by type, tags, or content.

40+ adapters

Type detection picks the right one. Works with whatever you have installed.

Jupyter history

In notebooks, PLS traces the code that created your object. A tiny time machine.

result = pls.describe(df_filtered)
print(result.history)

# ['df = pd.read_csv("data.csv")',
#  'df_clean = df.dropna()',
#  'df_filtered = df_clean[df_clean["value"] > 100]']

How it works