Legend explains the legend of advanced visualization library Plotly

Public account: You and the cabin by: Peter Editor: Peter

Hello, I’m Peter

It has been a long time since I updated a Plotly related article. The topic I chose was “Playing with Plotly legend Settings”, which is a topic I’ve been thinking about writing about for a long time. The main content of this article is:

Official website: plotly.com/python/lege…

Serialized article by Plotly

The Plotly article has been updated to article 16. Please read it. Recommend a few articles:

Import libraries and data

The data mainly used in this article is consumption data:

import pandas as pd
import numpy as np

import plotly_express as px
import plotly.graph_objects as go

# Consumption data

tips = px.data.tips()
tips.head()
Copy the code

Number of legend elements

It is mainly a legend composed of a single element and multiple elements.

A single element

fig = px.scatter(
    tips, # database
    x="total_bill".# xy
    y="tip",
    color="sex"   # Legend is distinguished by color
)

fig.show()
Copy the code

A legend composed of multiple elements

When multiple elements are distinguished at the same time, different shapes are generated to display data:

fig = px.scatter(
    tips,
    x="total_bill",
    y="tip",
    color="sex".# color and tag to distinguish
    symbol="smoker"
)

fig.show()
Copy the code

Change the legend name

With plotly_express, the labels parameter is used to change the legend name

fig = px.scatter(
    tips, 
    x="total_bill", 
    y="tip", 
    color="sex".# color and markup; Symbol mainly controls shape
    symbol="smoker".# Change the default legend name with labels
    labels={"sex": "Gender"."smoker": "Smokes"})

fig.show()
Copy the code

You can also generate a graph with more than one cut graph:

fig = px.scatter(
    tips, 
    x="total_bill", 
    y="tip", 
    color="sex", 
    symbol="smoker", 
    facet_col="time".# The column element of the section diagram
    # change the legend name
    labels={"sex": "Gender"."smoker": "Smokes"})

fig.show()
Copy the code

Set the legend order

Plot by plotly_express

The legend order is set with the legend. Traceorder parameter

fig = px.bar(
    tips, # data frame
    x="day".# xy
    y="total_bill", 
    color="smoker".# color
    barmode="stack".# Bar chart pattern
    facet_col="sex".# The column element of the section diagram
    category_orders={"day": ["Thur"."Fri"."Sat"."Sun"].# Custom order
                     "smoker": ["Yes"."No"]."sex": ["Male"."Female"]})

fig.update_layout(legend_traceorder="reversed")  # set order

fig.show()
Copy the code

⚠️ defaults to the order in which the fields of the real legend appear in the raw data:

fig = px.bar(
    tips, # data frame
    x="day".# xy
    y="total_bill", 
    color="smoker".# color
    barmode="stack".# Bar chart pattern
    facet_col="sex".# The column element of the section diagram
)

fig.show()
Copy the code

Plot with Plotly. Graph_objects

We set the order primarily through metrics Legendrank. Note: Plotly must be V5.0 or later

pip install --upgrade plotly  # Upgrade version
Copy the code

If we don’t use the Legendrank parameter:

fig = go.Figure()

# Add 4 different track data
fig.add_trace(go.Bar(name="first", x=["a"."b"], y=[21.27]))
fig.add_trace(go.Bar(name="second", x=["a"."b"], y=[32.18]))
fig.add_trace(go.Bar(name="third", x=["a"."b"], y=[11.32]))
fig.add_trace(go.Bar(name="fourth", x=["a"."b"], y=[21.13]))

fig.show()
Copy the code

The order in which the legends appear above is the order of first, second, third and fourth, that is, the order in which track data is added.

Here we use the Legendrank parameter:

fig = go.Figure()

# Add 4 different track data
fig.add_trace(go.Bar(name="second".# 2
                     x=["a"."b"], 
                     y=[32.18],
                     legendrank=2))

fig.add_trace(go.Bar(name="third".# 3
                     x=["a"."b"],
                     y=[11.32],
                     legendrank=3))

fig.add_trace(go.Bar(name="first".# 1
                     x=["a"."b"], 
                     y=[21.27],
                     legendrank=1))

fig.add_trace(go.Bar(name="fourth".# 4
                     x=["a"."b"], 
                     y=[21.13],
                     legendrank=4))

fig.show()
Copy the code

The order in which trace is added varies from first to second. In fact, it still sorts according to legendrank parameters

Show and hide legends

All of the above graphics are shown as legends by default, we can also hide them:

# Hide legends

fig = px.scatter(
    tips,
    x="total_bill",
    y="tip",
    color="sex",  
    symbol="smoker"
)

fig.update_layout(showlegend=False)  # Hide the legend element

fig.show()
Copy the code

Legend location

Legends are on the right by default, and we can use arguments to change the legend position

# Default: Legend in the upper right corner

fig = px.scatter(
    tips,
    x="total_bill",
    y="tip",
    size="tip",
    color="sex".# color and tag to distinguish
    symbol="smoker"
)

fig.show()
Copy the code

Change the position of legend by setting parameters:

# default

fig = px.scatter(
    tips,
    x="total_bill",
    y="tip",
    size="tip",
    color="sex".# color and tag to distinguish
    symbol="smoker"
)

# change by position parameter
fig.update_layout(legend=dict(
    yanchor="top".# top of the Y-axis
    y=0.99,
    xanchor="left".The X-axis is to the left
    x=0.01
))

fig.show()
Copy the code

Legend level

The default legend is vertical display, with the parameter orientation to achieve horizontal display. This horizontal display is useful when there is a lot of categorization of legends.

This example uses GDP data from Plotly:

The default legend is vertical:

df = px.data.gapminder().query("year==2007")  # select 2007 data

fig = px.scatter(
    df,  # Data box and x and y axis data
    x="gdpPercap", 
    y="lifeExp", 
    color="continent".# color
    size="pop".# size
    size_max=45)  A maximum #

fig.update_layout(legend=dict(
Orientation ="h"
    yanchor="bottom".# set the position and distance of the xy axis respectively
    y=1.02,
    xanchor="right",
    x=1
))

fig.show()
Copy the code

Now let’s turn on the horizontal display legend:

# Enable horizontal display

fig = px.scatter(
    df, 
    x="gdpPercap", 
    y="lifeExp", 
    color="continent",
    size="pop", 
    size_max=45)

fig.update_layout(legend=dict(
    orientation="h".# Enable horizontal display
    yanchor="bottom",
    y=1.02,
    xanchor="right",
    x=1
))

fig.show()
Copy the code

Legend personalized display

fig = px.scatter(
    df, 
    x="gdpPercap", 
    y="lifeExp", 
    color="continent",
    size="pop", 
    size_max=45
)


fig.update_layout(
    legend=dict(
        x=82..# Set the legend position
        y=0,
        traceorder="reversed",  
        title_font_family="Times New Roman".# legend title font
        font=dict(  # legend font
            family="Courier",
            size=13,
            color="red"  # Color: red
        ),
        bgcolor="LightSteelBlue".# Legend background color
        bordercolor="Black".# Legend border color and width
        borderwidth=2
    )
)

fig.show()
Copy the code

Graph Objects sets the legend

The following examples use plotly. Graph_objects to illustrate the Settings:

• The name of the legend

The name of the legend

fig = go.Figure()

fig.add_trace(go.Scatter(
    x=[1.2.3.4.5],
    y=[1.2.3.4.5],
    name="Figure 1"  # legend name
))

fig.add_trace(go.Scatter(
    x=[1.2.3.4.5],
    y=[5.4.3.2.1],
    name=Figure 2 ""  # legend name
))

fig.show()
Copy the code

Legend title

fig = go.Figure()

fig.add_trace(go.Scatter(
    x=[1.2.3.4.5],
    y=[1.2.3.4.5],
    name="Figure 1"
))

fig.add_trace(go.Scatter(
    x=[1.2.3.4.5],
    y=[5.4.3.2.1],
    name=Figure 2 ""
))

# Key parameter: Set the legend title
fig.update_layout(legend_title_text='Legend heading')

fig.show()
Copy the code

Show or hide legends

import plotly.graph_objects as go

fig = go.Figure()

fig.add_trace(go.Scatter(
    x=list(range(5)),
    y=list(range(5)),
    showlegend=False  # do not display a legend for this trajectory
))


fig.add_trace(go.Scatter(
    x=list(range(5)),
    y=list(range(5[: : -))1],
))

fig.update_layout(showlegend=True) # Overall legend display

fig.show()
Copy the code

Sets the legend element size

import plotly.graph_objects as go

fig = go.Figure()

fig.add_trace(go.Scatter(
    x=list(range(5)),
    y=list(range(5)),
    mode='markers+lines'.# Mode selection
    marker={'size':10}  # tag size
))

fig.add_trace(go.Scatter(
    x=list(range(5)),
    y=[5.4.3.2.1],
    mode='lines+markers',
    marker={'size':50}
))

fig.update_layout(legend= {'itemsizing': 'constant'})  # Trace and constant Settings

fig.show()
Copy the code

When Itemsizing is set to constant:

When Itemsizing is trace: