[Python] What is the purpose of the output parameter in the server function of a shiny app?

In R, the output is used to render outputs by assigning the result of a render function to some element of the output object. It does not seem to work this way in python.

Some example python code:

def server(input: Inputs, output: Outputs, session: Session) -> None:
@render.plot
def hist() -> None:
hue = "species" if input.species() else None
sns.kdeplot(df, x=input.var(), hue=hue)
if input.show_rug():
sns.rugplot(df, x=input.var(), hue=hue, color="black", alpha=0.25)


In python, rendering is controlled with decorators and having a function name matching the name of a component defined in the UI. So when is the output parameter used for anything?

Full working sample


import seaborn as sns

# Import data from shared.py
from shared import df
from shiny import App, render, ui

app_ui = ui.page_sidebar(
ui.sidebar(
ui.input_select(
"var", "Select variable", choices=["bill_length_mm", "body_mass_g"]
),
ui.input_switch("species", "Group by species", value=True),
ui.input_switch("show_rug", "Show Rug", value=True),
),
ui.output_plot("hist"),
title="Hello sidebar!",
)


def server(input, output, session):
@render.plot
def hist():
hue = "species" if input.species() else None
sns.kdeplot(df, x=input.var(), hue=hue)
if input.show_rug():
sns.rugplot(df, x=input.var(), hue=hue, color="black", alpha=0.25)


app = App(app_ui, server)

Continue reading...