如何在Plotly Treemap中强制指定Timepoint的展示顺序？

使用Plotly Express制作Treemap时，Timepoint包含「Time 1」「Time 2」「Time 3」三个选项，但由于Time 2的数据量（访问量、仪器类别及仪器数量）少于Time 3，导致图中Time 3显示在Time 2之前，希望调整为Time 1→Time 2→Time 3的线性顺序。

原代码：

import plotly.express as px
import numpy as np
import pandas as pd

df = pd.read_csv("testdata.csv") 
df["all"] = "all" 
print(df)
fig = px.treemap(df, 
                 path=['Timepoint', 'Visit', 'Instrument Category', 'Instrument'],
                 maxdepth =2,
                 branchvalues = "total")
fig.update_traces(root_color="lightgrey")
fig.update_layout(margin = dict(t=50, l=25, r=25, b=25))
fig.show()
fig.write_html("treemap.html", auto_open=True)

曾尝试的无效操作：

• 误以为最新版Plotly不支持category_orders参数，添加后未生效（实际支持）
• 创建排序键但未正确应用：

time_order = {"Time 1": 1, "Time 2": 2, "Time 3": 3}
df["Timepoint_sort"] = df["Timepoint"].map(time_order)

方法1：直接用category_orders指定顺序 #

px.treemap支持category_orders参数，只需明确指定Timepoint的期望顺序即可：

import plotly.express as px
import numpy as np
import pandas as pd

df = pd.read_csv("testdata.csv") 
df["all"] = "all" 

fig = px.treemap(df, 
                 path=['Timepoint', 'Visit', 'Instrument Category', 'Instrument'],
                 maxdepth=2,
                 branchvalues="total",
                 # 指定Timepoint的线性顺序
                 category_orders={"Timepoint": ["Time 1", "Time 2", "Time 3"]})
fig.update_traces(root_color="lightgrey")
fig.update_layout(margin=dict(t=50, l=25, r=25, b=25))
fig.show()
fig.write_html("treemap.html", auto_open=True)

方法2：将Timepoint转为Pandas有序分类 #

通过修改数据类型，让Plotly自动遵循预设的分类顺序：

import plotly.express as px
import numpy as np
import pandas as pd

df = pd.read_csv("testdata.csv") 
df["all"] = "all" 

# 把Timepoint设置为有序分类
time_sequence = ["Time 1", "Time 2", "Time 3"]
df["Timepoint"] = pd.Categorical(df["Timepoint"], categories=time_sequence, ordered=True)

fig = px.treemap(df, 
                 path=['Timepoint', 'Visit', 'Instrument Category', 'Instrument'],
                 maxdepth=2,
                 branchvalues="total")
fig.update_traces(root_color="lightgrey")
fig.update_layout(margin=dict(t=50, l=25, r=25, b=25))
fig.show()
fig.write_html("treemap.html", auto_open=True)

补充说明 #

• 方法1适合临时调整顺序，直接在绘图时指定；方法2适合需要多次复用同一顺序的场景，修改一次数据类型即可。
• 之前的排序键方法无效是因为未将排序逻辑关联到绘图流程中，上述两种方法都是直接向Plotly传递明确的顺序规则。

内容的提问来源于stack exchange，提问作者Jane