如何更改Plotly sublopts的大小和间距？

Question

如何操纵行之间的间距（例如增加前两行和最后两行之间的间距）和图中图表的大小（例如使饼图更大）？

在我的示例中，我尝试使用plotly 的 FigureWidget 和 make_subplots 可视化来自

Telco Customer Churn dataset

（

download 176kB

）的多个数据列。

此代码循环遍历 8 列，为每列添加 1 个饼图和 1 个条形图。

import pandas as pd
import plotly.graph_objects as go
from plotly.subplots import make_subplots

# Read data
df = pd.read_csv("./WA_Fn-UseC_-Telco-Customer-Churn.csv")
df['SeniorCitizen'] = df['SeniorCitizen'].map({0: 'No', 1: 'Yes'})

# Define subplot titles and data columns
data_cols = ['PhoneService' ,'MultipleLines' ,'InternetService' ,'OnlineBackup' ,'DeviceProtection' ,'TechSupport' ,'StreamingTV' ,'StreamingMovies']
titles = ['Phone Service' ,'Multiple Lines' ,'Internet Service' ,'Online Backup' ,'Device Protection' ,'Tech Support' ,'Streaming TV' ,'Streaming Movies']

fig = go.FigureWidget(make_subplots(rows=4, cols=4, specs=[
                                                        [{'type':'domain'}, {'type':'domain'}, {'type':'domain'}, {'type':'domain'}],
                                                        [{'type':'xy'}, {'type':'xy'}, {'type':'xy'}, {'type':'xy'}],
                                                        [{'type':'domain'}, {'type':'domain'}, {'type':'domain'}, {'type':'domain'}],
                                                        [{'type':'xy'}, {'type':'xy'}, {'type':'xy'}, {'type':'xy'}]]))

row, col = 1, 0
for i, (title, data_col) in enumerate(zip(titles, data_cols)):
    row, col = divmod(i, 4)
    row = row * 2
    
    # Get value counts for pie chart
    value_counts = df[data_col].value_counts()
    # Create pie chart trace and add to subplot
    pie_chart = go.Pie(labels=value_counts.index, values=value_counts.to_numpy(), name=title, title=title)
    fig.add_trace(pie_chart, row=row+1, col=col+1)
 
    # get churn rates
    churn_counts = df.groupby([data_col, 'Churn'])['Churn'].count().unstack()
    # Create stacked bar charts
    t1 = go.Bar(name='Churn (yes)', x=churn_counts['Yes'].index, y=churn_counts['Yes'])
    t2 = go.Bar(name='Churn (no)', x=churn_counts['No'].index, y=churn_counts['No'], marker_color='indianred')
    fig.add_trace(t1, row=row+2, col=col+1)
    fig.add_trace(t2, row=row+2, col=col+1)


fig.update_layout(title="Distribution of Customer Services", barmode='stack', showlegend=False)
fig.show()

---

编辑：将列数减少到两列也无法解决此问题。这是大屏幕上的图表：

import pandas as pd
import plotly.graph_objects as go
from plotly.subplots import make_subplots

# Read data
df = pd.read_csv("./WA_Fn-UseC_-Telco-Customer-Churn.csv")
df['SeniorCitizen'] = df['SeniorCitizen'].map({0: 'No', 1: 'Yes'})

# Define subplot titles and data columns
data_cols = ['PhoneService' ,'MultipleLines' ,'InternetService' ,'OnlineBackup' ,'DeviceProtection' ,'TechSupport' ,'StreamingTV' ,'StreamingMovies']
titles = ['Phone Service' ,'Multiple Lines' ,'Internet Service' ,'Online Backup' ,'Device Protection' ,'Tech Support' ,'Streaming TV' ,'Streaming Movies']

fig = go.FigureWidget(make_subplots(rows=8, cols=2, specs=[
                                                        [{'type':'domain'}, {'type':'domain'}],
                                                        [{'type':'xy'}, {'type':'xy'}],
                                                        [{'type':'domain'}, {'type':'domain'}],
                                                        [{'type':'xy'}, {'type':'xy'}],
                                                        [{'type':'domain'}, {'type':'domain'}],
                                                        [{'type':'xy'}, {'type':'xy'}],
                                                        [{'type':'domain'}, {'type':'domain'}],
                                                        [{'type':'xy'}, {'type':'xy'}]]))

row, col = 1, 0
for i, (title, data_col) in enumerate(zip(titles, data_cols)):
    row, col = divmod(i, 2)
    row = row * 2

    # Get value counts for pie chart
    value_counts = df[data_col].value_counts()
    # Create pie chart trace and add to subplot
    pie_chart = go.Pie(labels=value_counts.index, values=value_counts.to_numpy(), name=title, title=title)
    fig.add_trace(pie_chart, row=row+1, col=col+1)

    # get churn rates
    churn_counts = df.groupby([data_col, 'Churn'])['Churn'].count().unstack()
    # Create stacked bar charts
    t1 = go.Bar(name='Churn (yes)', x=churn_counts['Yes'].index, y=churn_counts['Yes'])
    t2 = go.Bar(name='Churn (no)', x=churn_counts['No'].index, y=churn_counts['No'], marker_color='indianred')
    fig.add_trace(t1, row=row+2, col=col+1)
    fig.add_trace(t2, row=row+2, col=col+1)


fig.update_layout(title="Distribution of Customer Services", barmode='stack', showlegend=False)
fig.show()

Answer 1

行和列之间的间距可以分别使用

vertical_spacing

和

horizontal_spacing

参数进行调整。两者都可以有 0 到

1/(size -1)

之间的浮点值。需要进行一些修改才能找到适合您的正确值。要最大化图中图表的大小，您可以更新

margin

调用中的

fig.update_layout

参数。所有四个边都在字典中用以下键引用：

t

、

b

、

l

和

r

。

下面的代码是使用所有这些参数对您的代码进行的更改。

import pandas as pd
import plotly.graph_objects as go
from plotly.subplots import make_subplots

# Read data
df = pd.read_csv(".../WA_Fn-UseC_-Telco-Customer-Churn.csv")
df['SeniorCitizen'] = df['SeniorCitizen'].map({0: 'No', 1: 'Yes'})

# Define subplot titles and data columns
data_cols = ['PhoneService' ,'MultipleLines' ,'InternetService' ,'OnlineBackup' ,'DeviceProtection' ,'TechSupport' ,'StreamingTV' ,'StreamingMovies']
titles = ['Phone Service' ,'Multiple Lines' ,'Internet Service' ,'Online Backup' ,'Device Protection' ,'Tech Support' ,'Streaming TV' ,'Streaming Movies']

hor_space = 0.02
ver_space = 0.02

fig = go.FigureWidget(make_subplots(rows=4,
                                    cols=4,
                                    specs=[[{'type':'domain'}, {'type':'domain'}, {'type':'domain'}, {'type':'domain'}],
                                           [{'type':'xy'}, {'type':'xy'}, {'type':'xy'}, {'type':'xy'}],
                                           [{'type':'domain'}, {'type':'domain'}, {'type':'domain'}, {'type':'domain'}],
                                           [{'type':'xy'}, {'type':'xy'}, {'type':'xy'}, {'type':'xy'}]
                                           ],
                                    horizontal_spacing=hor_space, # in range 0 to 1/(cols-1)
                                    vertical_spacing=ver_space # in range 0 to 1/(rows-1)
                                    )
                      )

row, col = 1, 0
for i, (title, data_col) in enumerate(zip(titles, data_cols)):
    row, col = divmod(i, 4)
    row = row * 2
    
    # Get value counts for pie chart
    value_counts = df[data_col].value_counts()
    # Create pie chart trace and add to subplot
    pie_chart = go.Pie(labels=value_counts.index, values=value_counts.to_numpy(), name=title, title=title)
    fig.add_trace(pie_chart, row=row+1, col=col+1)
 
    # get churn rates
    churn_counts = df.groupby([data_col, 'Churn'])['Churn'].count().unstack()
    # Create stacked bar charts
    t1 = go.Bar(name='Churn (yes)', x=churn_counts['Yes'].index, y=churn_counts['Yes'])
    t2 = go.Bar(name='Churn (no)', x=churn_counts['No'].index, y=churn_counts['No'], marker_color='indianred')
    fig.add_trace(t1, row=row+2, col=col+1)
    fig.add_trace(t2, row=row+2, col=col+1)


fig.update_layout(title="Distribution of Customer Services",
                  barmode='stack',
                  showlegend=False,
                   margin={"l":25,
                           "r":25,
                           "t":25,
                           "b":25}
                  )
fig.show()

plotly 文档提供的一些背景：

• 次要情节
• 调整图表大小