Pretty Little Summary
Automatic structured summaries of Python objects — DataFrames, arrays, models, plots, and more.
Install
pip install pretty-little-summary
Optional adapters are enabled automatically when their libraries are installed.
No install? Try it in the browser: live playground (runs via Pyodide, nothing uploaded).
Features
- Single function API:
pls.describe(obj) - 40+ adapters across data, viz, and ML libraries
- Works with built-ins out of the box (no required deps)
- Jupyter/IPython history capture for better context
- Deterministic, bounded question-focused tabular views with pluggable relevance scorers
Quick Start
import pretty_little_summary as pls
import pandas as pd
df = pd.DataFrame({
"product": ["Widget", "Gadget", "Doohickey"],
"price": [19.99, 29.99, 39.99],
"quantity": [100, 50, 75]
})
result = pls.describe(df)
print(result.content)
print(result.meta)
# Rank all columns for a question without changing the full profile.
focused = pls.focus_profile(
result,
"How does price relate to quantity?",
max_columns=8,
max_chars=4000,
)
print(focused.content)
Built-in Types
import pretty_little_summary as pls
print(pls.describe([1, 2, 3]).content)
print(pls.describe({"name": "Alice", "age": 30}).content)
NumPy Arrays
import numpy as np
import pretty_little_summary as pls
arr = np.random.rand(100, 50)
result = pls.describe(arr)
print(result.content)
Pandas DataFrames
import pandas as pd
import pretty_little_summary as pls
df = pd.read_csv("data.csv")
result = pls.describe(df)
print(result.content)
Files and dynamic imports
File paths are detected through a shared capability registry using extensions
and, where available, magic bytes. In native Python, dynamic_imports=True
imports matching optional libraries that are already installed; it never runs a
package manager or accesses the network:
import pretty_little_summary as pls
result = pls.describe("measurements.h5", dynamic_imports=True)
print(result.content)
# Hosts that manage packages themselves can inspect the same requirements:
print(pls.requirements_for_path("measurements.h5"))
The browser playground uses these requirements to install missing Pyodide or micropip packages automatically, so there is no package-tier selector.
Matplotlib Figures
import matplotlib.pyplot as plt
import pretty_little_summary as pls
fig, ax = plt.subplots()
ax.plot([1, 2, 3], [4, 5, 6])
result = pls.describe(fig)
print(result.content)
History Tracking (Jupyter/IPython)
When running inside Jupyter, pretty_little_summary can capture recent code history that created your object:
import pandas as pd
import pretty_little_summary as pls
df = pd.read_csv("data.csv")
df_clean = df.dropna()
result = pls.describe(df_clean)
print(result.history)
Docs site
The docs (docs/index.html, docs/playground.html, docs/eval-report.html)
are a static site with no build step. To run it locally:
cd docs && python3 -m http.server 8000
Then open http://localhost:8000. Notes:
playground.htmlloads Pyodide from a CDN, so it needs internet access even when served locally.playground.htmlinstallspretty-little-summaryfromdocs/dist/*.whlso it always matches this commit'ssrc/instead of a possibly-stale PyPI release. That wheel isn't committed — CI builds it on every docs deploy viapython -m build --wheel -o docs/dist. Build it yourself before serving locally, or the playground falls back to installing from PyPI:.venv/bin/python -m build --wheel -o docs/dist
eval-report.htmlisn't committed — generate it first with:.venv/bin/python -m evals.runner fetch .venv/bin/python -m evals.runner run .venv/bin/python -m evals.viewer --out docs/eval-report.html
(CI does this automatically on every docs deploy.) Its markup lives in the committeddocs/eval-report-template.html;evals/viewer.pyonly injects run data into it.- The playground's curated/gallery example code comes from
docs/examples/*.txt, regenerated via:.venv/bin/python scripts/export_examples.py
Troubleshooting
ModuleNotFoundError: No module named 'pretty_little_summary'
- Ensure you installed the package in the current environment.
- Restart your kernel or interpreter.
Missing optional libraries
If an adapter isn’t available, install its library:
pip install pandas numpy matplotlib
Or install all optional dependencies:
pip install pretty-little-summary[all]
Metadata
Release files for pretty-little-summary 0.3.1
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | Provenance | Actions |
|---|---|---|---|---|
| pretty_little_summary-0.3.1.tar.gz | 197.6 kB | Without provenance | Details for pretty_little_summary-0.3.1.tar.gz |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Actions Reset |
|---|---|---|---|---|
| pretty_little_summary-0.3.1-py3-none-any.whl | Python 3 | none | any | Details for pretty_little_summary-0.3.1-py3-none-any.whl |
Total release size: 307.2 kB
Release files / pretty_little_summary-0.3.1.tar.gz
| Download URL | pretty_little_summary-0.3.1.tar.gz |
|---|---|
| Size | 197.6 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
37d8ed10d92fe98eab4f19b11124e651b86870d0d5efae62b8f5ca3521ac6fce
|
|
BLAKE2b-256 checksum How to use checksums |
a270d419f8cc88112b8d924bef5c26a8b6aa4f2eac9b5e55b7e23910ba3f852f
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.12.14
|
Release files / pretty_little_summary-0.3.1-py3-none-any.whl
| Download URL | pretty_little_summary-0.3.1-py3-none-any.whl |
|---|---|
| Size | 109.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
f93f1ffc2639e8e16c670059dd2f6ff4d85110ee1c998da3e0808fa6ca61600f
|
|
BLAKE2b-256 checksum How to use checksums |
258d1c41b3187a49482c0ca5f8a75f12a9d391515ba31b7b6548ba76d05fd687
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/7.0.0 CPython/3.12.14
|
Security for pretty-little-summary 0.3.1
Provenance
Provenance describes where a file came from. On PyPI, provenance is shared via attestations, which provide a verifiable record of the build or publishing details. View details, limitations and caveats.
No provenance found
None of the files in this release were uploaded with provenance metadata.
A note on package security
PyPI is a repository for distributing packages, not a service for auditing them. While third-party security researchers and our staff do their best to detect and remove malicious packages, PyPI cannot guarantee the quality, security or suitability of any given package.
To learn more about protecting yourself from malicious actors, visit our package security documentation.