A reactive notebook for Python — run reproducible experiments, query with SQL, execute as a script, deploy as an app, and version with git. Stored as pure Python. All in a modern, AI-native editor.
Python artificial-intelligencedagdata-sciencedata-visualizationdataflowdeveloper-toolsmachine-learningnotebookspipelinepythonreactivesqlweb-app
# /// script
# requires-python = ">=3.12"
# dependencies = [
# "altair==5.5.0",
# "polars==1.17.1",
# "marimo",
# "quak==0.2.1",
# "pandas",
# ]
# ///
import marimo
__generated_with = "0.19.7"
app = marimo.App(width="medium")
@app.cell
def _():
import marimo as mo
import polars as pl
import quak
return mo, pl, quak
@app.cell
def _(mo, pl, quak):
df = pl.read_csv(
"https://raw.githubusercontent.com/uwdata/mosaic/main/data/athletes.csv"
)
widget = mo.ui.anywidget(quak.Widget(df))
widget
return (widget,)
@app.cell
def _(grouped_selection, mo):
import altair as alt
mo.ui.altair_chart(
alt.Chart(grouped_selection)
.mark_bar()
.encode(y=alt.Y("nationality:N").sort("-x"), x="count:Q")
.transform_window(
rank="rank(count)", sort=[alt.SortField("count", order="descending")]
)
.transform_filter(alt.datum.rank < 10)
.properties(height=400)
)
return
@app.cell
def _(widget):
selection = widget.data().df()
return (selection,)
@app.cell
def _(selection):
selection["count"] = 1
grouped_selection = (
selection[["nationality", "count"]]
.groupby(["nationality"])
.agg("count")
.reset_index()
)
return (grouped_selection,)
if __name__ == "__main__":
app.run()