EDA on a server¶

This example shows how to inspect data and plots when working on a server or terminal-based environment.

The example publishes:

a DataFrame
a summary object
a matplotlib plot

Install dependencies¶

pip install plotsrv polars matplotlib

or:

uv add plotsrv polars matplotlib

Create the script¶

Create eda_example.py:

eda_example.py

import matplotlib.pyplot as plt
import polars as pl
import plotsrv as ps


HOST = "127.0.0.1"
PORT = 8000


def main() -> None:
    df = pl.DataFrame(
        {
            "month": ["Jan", "Feb", "Mar", "Apr", "May", "Jun"],
            "bookings": [120, 150, 170, 160, 210, 240],
            "returns": [115, 148, 165, 155, 205, 235],
        }
    )

    summary = {
        "rows": df.height,
        "total_bookings": df["bookings"].sum(),
        "total_returns": df["returns"].sum(),
        "max_bookings": df["bookings"].max(),
        "max_returns": df["returns"].max(),
    }

    fig, ax = plt.subplots()
    ax.plot(df["month"], df["bookings"], marker="o", label="bookings")
    ax.plot(df["month"], df["returns"], marker="o", label="returns")
    ax.set_title("Bookings and returns")
    ax.set_xlabel("Month")
    ax.set_ylabel("Count")
    ax.legend()

    ps.publish_view(
        df,
        label="monthly data",
        section="eda",
        host=HOST,
        port=PORT,
    )

    ps.publish_view(
        summary,
        label="summary",
        section="eda",
        host=HOST,
        port=PORT,
    )

    ps.publish_view(
        fig,
        label="plot",
        section="eda",
        host=HOST,
        port=PORT,
    )


if __name__ == "__main__":
    main()

Start plotsrv¶

In one terminal:

plotsrv run eda_example.py --host 127.0.0.1 --port 8000

Open:

http://127.0.0.1:8000

Run the script¶

In another terminal:

python eda_example.py

The UI should show:

the source data as a table
summary values as JSON
the matplotlib plot as an image view

SSH workflow¶

For remote server work, use port forwarding from the local machine:

ssh -L 8000:127.0.0.1:8000 user@server

Then run plotsrv on the server:

plotsrv run eda_example.py --host 127.0.0.1 --port 8000

Open locally:

http://127.0.0.1:8000

This avoids exposing plotsrv directly to the network.

Attached version¶

For a quick local check, use launch_server=True instead of host and port:

ps.publish_view(
    df,
    label="monthly data",
    section="eda",
    launch_server=True,
)

Keep the process open at the end of the script:

input("Open http://127.0.0.1:8000, then press Enter to stop...")

What this example shows¶

plotsrv is useful when a script produces objects that are better inspected visually than printed:

tables are easier to scan in a browser
JSON summaries are easier to expand and inspect
plots can be viewed on a headless server
repeated script runs update the same UI