Plotly's reactive app framework - see dashboards page. Shares figure model with Express/GO.

px.scatter(..., animation_frame="year") for small datasets. Avoid animation on huge frames.

Plotly

Q: Can I use Plotly in Jupyter?

fig.show() renders inline with nbformat support. Prefer HTML export for sharing outside notebook.

Q: Does Plotly work with Polars?

Pass pl.DataFrame where supported or .to_pandas() first.

Q: How do I embed in Flask?

Return fig.to_html(full_html=False) in templates. Load plotly.js once in base layout.

Plotly builds interactive, web-ready charts with hover, zoom, and pan - export HTML or embed in Dash apps.

Recipe

Quick-reference recipe card - copy-paste ready.

import plotly.express as px
import pandas as pd
 
df = pd.read_csv("sales.csv", parse_dates=["ordered_at"])
fig = px.line(df, x="ordered_at", y="revenue", color="region", title="Revenue trend")
fig.update_layout(hovermode="x unified")
fig.write_html("revenue.html", include_plotlyjs="cdn")

When to reach for this:

Dashboards and stakeholder self-serve exploration
Dense scatter needing hover detail
Linked brushing across facets (Dash)
Embedding charts in internal web tools

Working Example

import plotly.express as px
import plotly.graph_objects as go
import pandas as pd
import numpy as np
 
rng = np.random.default_rng(42)
df = pd.DataFrame(
    {
        "region": np.repeat(["East", "West"], 100),
        "month": np.tile(np.arange(1, 13).repeat(8)[:100], 2),
        "revenue": rng.integers(80, 200, 200),
        "units": rng.integers(5, 50, 200),
    }
)
 
# Express scatter with custom hover
fig_scatter = px.scatter(
    df,
    x="units",
    y="revenue",
    color="region",
    hover_data=["month"],
    opacity=0.7,
    title="Units vs revenue",
)
fig_scatter.update_traces(marker=dict(size=8))
 
# Graph objects for dual-axis combo
monthly = df.groupby(["month", "region"], observed=True)["revenue"].sum().reset_index()
fig_combo = go.Figure()
for region, sub in monthly.groupby("region"):
    fig_combo.add_trace(
        go.Scatter(x=sub["month"], y=sub["revenue"], mode="lines+markers", name=region)
    )
fig_combo.update_layout(
    xaxis_title="Month",
    yaxis_title="Revenue",
    legend_title="Region",
    template="plotly_white",
)
 
fig_scatter.write_html("scatter.html", include_plotlyjs="cdn")
fig_combo.write_html("combo.html", include_plotlyjs="cdn")

What this demonstrates:

Plotly Express for rapid scatter with hover_data
Graph Objects for multi-trace figures
template for consistent light theme
CDN-hosted JS for smaller HTML files

Deep Dive

How It Works

Figures are JSON-serializable dicts rendered by plotly.js in the browser.
Express wraps graph_objects with sensible defaults from DataFrame columns.
Traces hold data; layout controls axes, legends, and annotations.
Dash builds reactive apps sharing the same figure model.

Express vs Graph Objects

Layer	Best for
`plotly.express`	Standard charts from tidy data
`plotly.graph_objects`	Custom layouts, dual axes, annotations

Python Notes

import plotly.io as pio
 
# Static image export needs kaleido
pio.write_image(fig, "chart.png", scale=2)

Gotchas

Plotting millions of points - browser hangs. Fix: aggregate, hexbin, or datashader first.
Huge HTML from inline JS - include_plotlyjs=True bloats files. Fix: use "cdn" or "directory".
Categorical x treated as numeric - lines connect wrongly. Fix: astype(str) or category_orders.
Timezone-naive datetimes - hover shows wrong local offset. Fix: UTC at source, format in hovertemplate.
Kaleido missing for PNG - write_image fails in CI. Fix: uv pip install kaleido in export jobs.

Alternatives

Alternative	Use When	Don't Use When
matplotlib	Print/PDF static pipelines	Need hover tooltips
Altair	Vega-Lite declarative specs	Need 3-D WebGL traces
Bokeh	Large streaming datasets	Team already on Plotly/Dash
Observable Plot	Web-native grammar	Python-only batch jobs

FAQs

Express or graph_objects?

Start Express; drop to GO for dual axes and custom annotations.
GO is verbose but unlimited.

How do I customize hover text?

fig.update_traces(hovertemplate="Revenue: %{y:$,.0f}<extra></extra>")

Can I use Plotly in Jupyter?

fig.show() renders inline with nbformat support.
Prefer HTML export for sharing outside notebook.

How do I fix legend order?

fig.update_layout(legend_traceorder="reversed")

Or sort traces before adding.

How do I add horizontal reference line?

fig.add_hline(y=150, line_dash="dash", annotation_text="target")

Does Plotly work with Polars?

Pass pl.DataFrame where supported or .to_pandas() first.

How do I dark theme?

fig.update_layout(template="plotly_dark")

How do I embed in Flask?

Return fig.to_html(full_html=False) in templates.
Load plotly.js once in base layout.

What is Dash?

Plotly's reactive app framework - see dashboards page.
Shares figure model with Express/GO.

How do I animate?

px.scatter(..., animation_frame="year") for small datasets.
Avoid animation on huge frames.

Dashboards (Streamlit / Dash / Gradio) - Dash apps
Plotting Large Data - downsampling
Altair - declarative alternative
Visualization Basics - when interactivity helps

Stack versions: This page was written for Python 3.14.0 (stable 3.14, maintenance 3.13), FastAPI 0.115+, Django 5.2, Flask 3.1, Pydantic 2, PyTorch 2.6+, pandas 2.2+, Polars 1.x, ruff 0.9+, and uv 0.6+.

Recipe

Quick-reference recipe card - copy-paste ready.

import plotly.express as px
import pandas as pd
 
df = pd.read_csv("sales.csv", parse_dates=["ordered_at"])
fig = px.line(df, x="ordered_at", y="revenue", color="region", title="Revenue trend")
fig.update_layout(hovermode="x unified")
fig.write_html("revenue.html", include_plotlyjs="cdn")

When to reach for this:

Dashboards and stakeholder self-serve exploration
Dense scatter needing hover detail
Linked brushing across facets (Dash)
Embedding charts in internal web tools

Working Example

import plotly.express as px
import plotly.graph_objects as go
import pandas as pd
import numpy as np
 
rng = np.random.default_rng(42)
df = pd.DataFrame(
    {
        "region": np.repeat(["East", "West"], 100),
        "month": np.tile(np.arange(1, 13).repeat(8)[:100], 2),
        "revenue": rng.integers(80, 200, 200),
        "units": rng.integers(5, 50, 200),
    }
)
 
# Express scatter with custom hover
fig_scatter = px.scatter(
    df,
    x="units",
    y="revenue",
    color="region",
    hover_data=["month"],
    opacity=0.7,
    title="Units vs revenue",
)
fig_scatter.update_traces(marker=dict(size=8))
 
# Graph objects for dual-axis combo
monthly = df.groupby(["month", "region"], observed=True)["revenue"].sum().reset_index()
fig_combo = go.Figure()
for region, sub in monthly.groupby("region"):
    fig_combo.add_trace(
        go.Scatter(x=sub["month"], y=sub["revenue"], mode="lines+markers", name=region)
    )
fig_combo.update_layout(
    xaxis_title="Month",
    yaxis_title="Revenue",
    legend_title="Region",
    template="plotly_white",
)
 
fig_scatter.write_html("scatter.html", include_plotlyjs="cdn")
fig_combo.write_html("combo.html", include_plotlyjs="cdn")

What this demonstrates:

Plotly Express for rapid scatter with hover_data
Graph Objects for multi-trace figures
template for consistent light theme
CDN-hosted JS for smaller HTML files

Deep Dive

How It Works

Figures are JSON-serializable dicts rendered by plotly.js in the browser.
Express wraps graph_objects with sensible defaults from DataFrame columns.
Traces hold data; layout controls axes, legends, and annotations.
Dash builds reactive apps sharing the same figure model.

Express vs Graph Objects

Layer	Best for
`plotly.express`	Standard charts from tidy data
`plotly.graph_objects`	Custom layouts, dual axes, annotations

Python Notes

import plotly.io as pio
 
# Static image export needs kaleido
pio.write_image(fig, "chart.png", scale=2)

Gotchas

Plotting millions of points - browser hangs. Fix: aggregate, hexbin, or datashader first.
Huge HTML from inline JS - include_plotlyjs=True bloats files. Fix: use "cdn" or "directory".
Categorical x treated as numeric - lines connect wrongly. Fix: astype(str) or category_orders.
Timezone-naive datetimes - hover shows wrong local offset. Fix: UTC at source, format in hovertemplate.
Kaleido missing for PNG - write_image fails in CI. Fix: uv pip install kaleido in export jobs.

Alternatives

Alternative	Use When	Don't Use When
matplotlib	Print/PDF static pipelines	Need hover tooltips
Altair	Vega-Lite declarative specs	Need 3-D WebGL traces
Bokeh	Large streaming datasets	Team already on Plotly/Dash
Observable Plot	Web-native grammar	Python-only batch jobs

FAQs

Express or graph_objects?

Start Express; drop to GO for dual axes and custom annotations.
GO is verbose but unlimited.

How do I customize hover text?

fig.update_traces(hovertemplate="Revenue: %{y:$,.0f}<extra></extra>")

Can I use Plotly in Jupyter?

fig.show() renders inline with nbformat support.
Prefer HTML export for sharing outside notebook.

How do I fix legend order?

fig.update_layout(legend_traceorder="reversed")

Or sort traces before adding.

How do I add horizontal reference line?

fig.add_hline(y=150, line_dash="dash", annotation_text="target")

Does Plotly work with Polars?

Pass pl.DataFrame where supported or .to_pandas() first.

How do I dark theme?

fig.update_layout(template="plotly_dark")

How do I embed in Flask?

Return fig.to_html(full_html=False) in templates.
Load plotly.js once in base layout.

What is Dash?

Plotly's reactive app framework - see dashboards page.
Shares figure model with Express/GO.

How do I animate?

px.scatter(..., animation_frame="year") for small datasets.
Avoid animation on huge frames.

Dashboards (Streamlit / Dash / Gradio) - Dash apps
Plotting Large Data - downsampling
Altair - declarative alternative
Visualization Basics - when interactivity helps

Stack versions: This page was written for Python 3.14.0 (stable 3.14, maintenance 3.13), FastAPI 0.115+, Django 5.2, Flask 3.1, Pydantic 2, PyTorch 2.6+, pandas 2.2+, Polars 1.x, ruff 0.9+, and uv 0.6+.

Plotly

Search Documentation

Plotly