如何在 plotly 快速子圖中添加餅圖和分組條形圖

Question

以下是使用 plotly 快速子圖獲取餅圖和分組條形圖的語法

import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots

specs = [[{'type':'pie'}, {"type": "bar"}]]
fig = make_subplots(rows=1, cols=2, specs=specs, shared_yaxes = True,subplot_titles=['Pie Chart',
                                                                                     'Grouped Bar Chart'])
                    
                    

fig.add_trace(go.Pie(
                            labels = df_pie['index'], 
                            values = df_pie['count'],
                            hole = 0.6,
                            marker_colors = ['#353837','#646665', '#8e9492', '#c9d1ce'],
                            ), 1, 1)  

fig.add_trace(go.Bar(
                        x = df_bar['Satisfaction'],
                        y = df_bar['count'],
                        base  =df_bar['Gender'],
                        ),1,2)

fig.update_layout(showlegend=False, 
                  title=dict(text="Visualization",
                             font=dict(
                                        family="Arial",
                                        size=20,
                                        color='#283747')
                    ))  

fig.show()

下面是我根據上面的代碼得到的結果，

我怎樣才能讓餅圖看起來像這樣

條形圖看起來像這樣

通過 plotly 表達子圖。

Answer 1

您可以通過將textinfo用於餅圖來實現此目的。 對於條形圖，您需要創建要在每個條形上顯示的文本，然后使用text來顯示它。 此外，當您正在尋找分組條形圖時，您將需要使用創建兩條軌跡，然后將其組合到子圖 - 1,2。 請注意， textposition=auto將 select 以正確的方式顯示文本。 在條形的情況下，由於長度，它已經移動了要垂直顯示的文本。

由於沒有提供數據，我創建了一些基本數據。 希望這就是您要找的。

import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots

specs = [[{'type':'pie'}, {"type": "bar"}]]
fig = make_subplots(rows=1, cols=2, specs=specs, shared_yaxes = True, subplot_titles=['Pie Chart', 'Grouped Bar Chart'])

##My data creation##                    
df_pie=pd.DataFrame({'index':[1,2,3,4], 'count':[442,459,289,280]})                    
df_bar=pd.DataFrame({'Satisfaction': ['Excellent', 'Excellent', 'Good', 'Good', 'Poor', 'Poor', 'Neutral', 'Neutral'], 'count': [442, 459, 289, 280, 442, 459, 289, 280], 'Gender': ['Male', 'Female', 'Male', 'Female', 'Male', 'Female', 'Male', 'Female']})                    

fig.add_trace(go.Pie(
                            labels = df_pie['index'], 
                            values = df_pie['count'],
                            hole = 0.6,
                            marker_colors = ['#353837','#646665', '#8e9492', '#c9d1ce'],
                            textinfo='percent+value',  ## ADD - display both
                            ), 1, 1)  

## New column to get percentage of row across all Male/Females
df_bar['percentage'] = df_bar.groupby(['Gender'])['count'].transform(lambda z: round(z / z.sum() * 100))

## New column - text of count and percentage - how you need the annotation shown
df_bar['Text']=df_bar['count'].astype(str) + ',' + df_bar['percentage'].astype(str)+'%'

## Create individual traces for Male and Female
trace1 = go.Bar(
    x=df_bar[df_bar.Gender == 'Male']['Satisfaction'],
    y=df_bar[df_bar.Gender == 'Male']['count'],
    name='Male',
    text=df_bar[df_bar.Gender == 'Male']['Text'], ## Display text
    textposition='auto',
)

trace2 = go.Bar(
    x=df_bar[df_bar.Gender == 'Female']['Satisfaction'],
    y=df_bar[df_bar.Gender == 'Female']['count'],
    name='Female',
    text=df_bar[df_bar.Gender == 'Male']['Text'], ## Display text
    textposition='auto'
)

fig.append_trace(trace1, 1,2) ## Add as first set of bars in second subplot
fig.append_trace(trace2,1,2) ## Add as second set of bars in second subplot

##Removed your original code
#fig.add_trace(go.Bar(
#                        x = df_bar['Satisfaction'],
#                        y = df_bar['count'],
#                        base  =df_bar['Gender'],
#                        ),1,2)

fig.update_layout(showlegend=False, 
                  title=dict(text="Visualization",
                             font=dict(
                                        family="Arial",
                                        size=20,
                                        color='#283747')
                    ))  

fig.show()