| jupyter | ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
|
Facet plots, also known as trellis plots or small multiples, are figures made up of multiple subplots which have the same set of axes, where each subplot shows a subset of the data. While it is straightforward to use plotly's
subplot capabilities to make such figures, it's far easier to use the built-in facet_row and facet_col arguments in the various Plotly Express functions.
Plotly Express is the easy-to-use, high-level interface to Plotly, which operates on a variety of types of data and produces easy-to-style figures.
import plotly.express as px
df = px.data.tips()
fig = px.scatter(df, x="total_bill", y="tip", color="smoker", facet_col="sex")
fig.show()import plotly.express as px
df = px.data.tips()
fig = px.bar(df, x="size", y="total_bill", color="sex", facet_row="smoker")
fig.show()When the facet dimension has a large number of unique values, it is possible to wrap columns using the facet_col_wrap argument.
import plotly.express as px
df = px.data.gapminder()
fig = px.scatter(df, x='gdpPercap', y='lifeExp', color='continent', size='pop',
facet_col='year', facet_col_wrap=4)
fig.show()import plotly.express as px
df = px.data.tips()
fig = px.histogram(df, x="total_bill", y="tip", color="sex", facet_row="time", facet_col="day",
category_orders={"day": ["Thur", "Fri", "Sat", "Sun"], "time": ["Lunch", "Dinner"]})
fig.show()new in version 4.13
import plotly.express as px
df = px.data.election()
df = df.melt(id_vars="district", value_vars=["Coderre", "Bergeron", "Joly"],
var_name="candidate", value_name="votes")
geojson = px.data.election_geojson()
fig = px.choropleth(df, geojson=geojson, color="votes", facet_col="candidate",
locations="district", featureidkey="properties.district",
projection="mercator"
)
fig.update_geos(fitbounds="locations", visible=False)
fig.show()introduced in plotly 4.12
It is possible to add labelled horizontal and vertical lines and rectangles to facet plots using .add_hline(), .add_vline(), .add_hrect() or .add_vrect(). The default row and col values are "all" but this can be overridden, as with the rectangle below, which only appears in the first column.
import plotly.express as px
df = px.data.stocks(indexed=True)
fig = px.line(df, facet_col="company", facet_col_wrap=2)
fig.add_hline(y=1, line_dash="dot",
annotation_text="Jan 1, 2018 baseline",
annotation_position="bottom right")
fig.add_vrect(x0="2018-09-24", x1="2018-12-18", col=1,
annotation_text="decline", annotation_position="top left",
fillcolor="green", opacity=0.25, line_width=0)
fig.show()introduced in plotly 4.12
The .add_trace() method can be used to add a copy of the same trace to each facet, for example an overall linear regression line as below. The legendgroup/showlegend pattern below is recommended to avoid having a separate legend item for each copy of the trace. Note that as of v5.2.1, there is a built-in option to add an overall trendline to all facets that uses this technique under the hood.
import plotly.express as px
df = px.data.tips()
fig = px.scatter(df, x="total_bill", y="tip", color='sex',
facet_col="day", facet_row="time")
import statsmodels.api as sm
import plotly.graph_objects as go
df = df.sort_values(by="total_bill")
model = sm.OLS(df["tip"], sm.add_constant(df["total_bill"])).fit()
#create the trace to be added to all facets
trace = go.Scatter(x=df["total_bill"], y=model.predict(),
line_color="black", name="overall OLS")
# give it a legend group and hide it from the legend
trace.update(legendgroup="trendline", showlegend=False)
# add it to all rows/cols, but not to empty subplots
fig.add_trace(trace, row="all", col="all", exclude_empty_subplots=True)
# set only the last trace added to appear in the legend
# `selector=-1` introduced in plotly v4.13
fig.update_traces(selector=-1, showlegend=True)
fig.show()By default, facet axes are linked together: zooming inside one of the facets will also zoom in the other facets. You can disable this behaviour when you use facet_row only, by disabling matches on the Y axes, or when using facet_col only, by disabling matches on the X axes. It is not recommended to use this approach when using facet_row and facet_col together, as in this case it becomes very hard to understand the labelling of axes and grid lines.
import plotly.express as px
df = px.data.tips()
fig = px.scatter(df, x="total_bill", y="tip", color='sex', facet_row="day")
fig.update_yaxes(matches=None)
fig.show()import plotly.express as px
df = px.data.tips()
fig = px.scatter(df, x="total_bill", y="tip", color='sex', facet_col="day")
fig.update_xaxes(matches=None)
fig.show()Since subplot figure titles are annotations, you can use the for_each_annotation function to customize them, for example to remove the equal-sign (=).
In the following example, we pass a lambda function to for_each_annotation in order to change the figure subplot titles from smoker=No and smoker=Yes to just No and Yes.
import plotly.express as px
fig = px.scatter(px.data.tips(), x="total_bill", y="tip", facet_col="smoker")
fig.for_each_annotation(lambda a: a.update(text=a.text.split("=")[-1]))
fig.show()By default, Plotly Express lays out categorical data in the order in which it appears in the underlying data. Every 2-d cartesian Plotly Express function also includes a category_orders keyword argument which can be used to control the order in which categorical axes are drawn, but beyond that can also control the order in which discrete colors appear in the legend, and the order in which facets are laid out.
import plotly.express as px
df = px.data.tips()
fig = px.bar(df, x="day", y="total_bill", color="smoker", barmode="group", facet_col="sex",
category_orders={"day": ["Thur", "Fri", "Sat", "Sun"],
"smoker": ["Yes", "No"],
"sex": ["Male", "Female"]})
fig.show()The facet_row_spacing and facet_col_spacing arguments can be used to control the spacing between rows and columns. These values are specified in fractions of the plotting area in paper coordinates and not in pixels, so they will grow or shrink with the width and height of the figure.
The defaults work well with 1-4 rows or columns at the default figure size with the default font size, but need to be reduced to around 0.01 for very large figures or figures with many rows or columns. Conversely, if activating tick labels on all facets, the spacing will need to be increased.
import plotly.express as px
df = px.data.gapminder().query("continent == 'Africa'")
fig = px.line(df, x="year", y="lifeExp", facet_col="country", facet_col_wrap=7,
facet_row_spacing=0.04, # default is 0.07 when facet_col_wrap is used
facet_col_spacing=0.04, # default is 0.03
height=600, width=800,
title="Life Expectancy in Africa")
fig.for_each_annotation(lambda a: a.update(text=a.text.split("=")[-1]))
fig.update_yaxes(showticklabels=True)
fig.show()Using facet_col from plotly.express let zoom and pan each facet to the same range implicitly. However, if the subplots are created with make_subplots, the axis needs to be updated with matches parameter to update all the subplots accordingly.
Zoom in one trace below, to see the other subplots zoomed to the same x-axis range. To pan all the subplots, click and drag from the center of x-axis to the side:
import plotly.graph_objects as go
from plotly.subplots import make_subplots
import numpy as np
N = 20
x = np.linspace(0, 1, N)
fig = make_subplots(1, 3)
for i in range(1, 4):
fig.add_trace(go.Scatter(x=x, y=np.random.random(N)), 1, i)
fig.update_xaxes(matches='x')
fig.show()