This release is a pre-release and may not be stable for production use.
dx
Smart DataFrame display for Jupyter, built for nteract.
dx upgrades how pandas and polars DataFrames render in a notebook. Instead of serializing megabytes of HTML into your output cells, dx hands the data to nteract's content-addressed blob store and renders it through a fast Arrow/parquet grid. Your .ipynb stays tiny, the cell stays snappy, and AI agents reading the notebook get a compact per-column summary — dtypes, ranges, distinct/top values, null counts — instead of raw bytes.
Install
# pandas
pip install --pre "dx[pandas]"
# polars
pip install --pre "dx[polars]"
# both
pip install --pre "dx[pandas,polars]"
Python 3.10+. Only pre-release wheels are being published while the library surface settles — the stable channel is frozen. See #2217. Most nteract users don't install dx directly: the kernel launcher calls dx.install() during bootstrap, so DataFrames render through the blob store inside the nteract desktop app automatically.
Use
import dx
dx.install()
import pandas as pd
df = pd.read_parquet("large-dataset.parquet")
df # rendered via nteract's sift grid — no base64 in your .ipynb
That's it. dx.install() is idempotent and automatically called by nteract's kernel bootstrap, so most nteract users never touch it directly. Calling it yourself is fine when you want the behavior in an environment nteract didn't configure for you (a standalone kernel, a test harness, etc.).
What you get
- Fast rendering. Large DataFrames stream through the blob store; the
.ipynbpayload stays small. - AI-friendly summaries. Every DataFrame ships a
text/llm+plaincolumn summary — dtypes, numeric ranges, string distinct/top values, null counts — so agents reason about the shape without materializing the whole table. - Visualization integration. Altair and Plotly are automatically switched to their nteract renderers for interactive output that works inside nteract's isolated iframe sandbox.
- Narwhals-aware. narwhals-wrapped DataFrames are unwrapped via
.to_native()and dispatched through the pandas/polars path. - Safe outside nteract. When no nteract runtime is reachable,
dx.install()is a no-op.import dxis safe from plain Python, vanilla Jupyter, scripts, CI.
Links
- Homepage: https://nteract.io
- Source & issues: https://github.com/nteract/desktop
- License: BSD-3-Clause
1 maintainer
Release files for dx 2.1.6a202605060946
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| dx-2.1.6a202605060946.tar.gz | 23.6 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| dx-2.1.6a202605060946-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 40.0 kB
Release files / dx-2.1.6a202605060946.tar.gz
| Download URL | dx-2.1.6a202605060946.tar.gz |
|---|---|
| Size | 23.6 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
d13126c0e998c3032433d9fe4dc56e58d27ebc65f3503fe2135436e103c5563a
|
|
BLAKE2b-256 checksum How to use checksums |
6d259597d14fe62bbd96b7d3bc0049d5d6c245969bf786d8df5a7fc7b7d4ee8e
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
uv/0.11.10 {"installer":{"name":"uv","version":"0.11.10","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
Release files / dx-2.1.6a202605060946-py3-none-any.whl
| Download URL | dx-2.1.6a202605060946-py3-none-any.whl |
|---|---|
| Size | 16.4 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
71fc29abc5c50b34d2b7880fa9e8850707eaa52ffe23c83ca20ec178ee84a8f5
|
|
BLAKE2b-256 checksum How to use checksums |
e76abbac7807c4e1cf7578140de932061524efb42c6510be293de91c1a2c2ae9
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
Yes |
| Uploaded via |
uv/0.11.10 {"installer":{"name":"uv","version":"0.11.10","subcommand":["publish"]},"python":null,"implementation":{"name":null,"version":null},"distro":{"name":"Ubuntu","version":"24.04","id":"noble","libc":null},"system":{"name":null,"release":null},"cpu":null,"openssl_version":null,"setuptools_version":null,"rustc_version":null,"ci":true}
|
