能否为Plotly堆叠柱状图添加颜色渐变？实现多维度颜色编码

我有一个基于Plotly Express的堆叠柱状图，X轴是20种期刊名称，每个期刊内按Access（访问类型）分为堆叠分段，且访问类型数据涵盖多个年份。请问能否同时通过颜色编码Access类型和Year（年份）？若可行，希望为每种颜色添加渐变来表示年份。

最小可复现代码（MWE） #

import pandas as pd
import plotly.express as px

d = {'Journal': ['Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One'],
     'Access': ['Green', 'Green', 'Closed', 'Green', 'Closed', 'Green', 'Closed'],
     'Year': [2020, 2021, 2021, 2022, 2022, 2023, 2023],
     'Count': [1, 1, 5, 6, 3, 2, 4]}
df = pd.DataFrame(data=d)

fig = px.bar(df, x='Journal', y='Count', color='Access', text_auto=True,
            category_orders={'Access': ["Closed", "Green"]},
            color_discrete_map={
                "Closed": "#AB63FA",
                "Green": "#2CA02C"},
           title = f'Journal Titles by Access status and Year' )

fig.show()

请问我能否按Year为绿色和紫色指定渐变或透明度级别，以区分颜色块内的不同数据？是否可以添加两种颜色编码？或者有更优的展示方式？

1. 给同一Access类型的不同年份添加颜色渐变/透明度 #

可以通过结合Access和Year生成新的分类字段，然后为每个分类设置深浅渐变的颜色，或者调整透明度来区分年份。

方法一：自定义渐变颜色映射

为每个Access类型定义颜色渐变范围，根据年份映射对应颜色：

import pandas as pd
import plotly.express as px
from plotly.colors import sample_colorscale

# 定义颜色渐变：年份越晚，颜色越深
closed_colors = sample_colorscale('purples', [0.3, 0.5, 0.7, 0.9])
green_colors = sample_colorscale('greens', [0.3, 0.5, 0.7, 0.9])

d = {'Journal': ['Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One'],
     'Access': ['Green', 'Green', 'Closed', 'Green', 'Closed', 'Green', 'Closed'],
     'Year': [2020, 2021, 2021, 2022, 2022, 2023, 2023],
     'Count': [1, 1, 5, 6, 3, 2, 4]}
df = pd.DataFrame(data=d)

# 生成Access+Year的组合字段
df['Access_Year'] = df['Access'] + '_' + df['Year'].astype(str)

# 构建颜色映射表
year_order = sorted(df['Year'].unique())
color_map = {}
for i, year in enumerate(year_order):
    color_map[f'Closed_{year}'] = closed_colors[i]
    color_map[f'Green_{year}'] = green_colors[i]

fig = px.bar(df, x='Journal', y='Count', color='Access_Year', text_auto=True,
             category_orders={
                 'Access_Year': [f'{access}_{year}' for access in ["Closed", "Green"] for year in year_order]
             },
             color_discrete_map=color_map,
             title='期刊按访问类型和年份统计' )

# 调整图例分组，提升可读性
fig.update_layout(
    legend=dict(
        tracegroupgap=20,
        itemsizing='constant'
    )
)

fig.show()

方法二：通过透明度区分年份

固定Access的基础颜色，仅按年份调整透明度（年份越晚透明度越高）：

import pandas as pd
import plotly.express as px

d = {'Journal': ['Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One', 'Journal One'],
     'Access': ['Green', 'Green', 'Closed', 'Green', 'Closed', 'Green', 'Closed'],
     'Year': [2020, 2021, 2021, 2022, 2022, 2023, 2023],
     'Count': [1, 1, 5, 6, 3, 2, 4]}
df = pd.DataFrame(data=d)

# 生成Access+Year的组合字段
df['Access_Year'] = df['Access'] + '_' + df['Year'].astype(str)

# 定义基础颜色，按年份调整透明度
base_colors = {"Closed": "#AB63FA", "Green": "#2CA02C"}
year_order = sorted(df['Year'].unique())
color_map = {}
for year in year_order:
    alpha = 0.5 + (year - min(year_order))/(max(year_order)-min(year_order))*0.4  # 透明度范围0.5-0.9
    for access in ["Closed", "Green"]:
        # 十六进制转RGBA格式
        hex_color = base_colors[access]
        r = int(hex_color[1:3], 16)
        g = int(hex_color[3:5], 16)
        b = int(hex_color[5:7], 16)
        color_map[f'{access}_{year}'] = f'rgba({r}, {g}, {b}, {alpha})'

fig = px.bar(df, x='Journal', y='Count', color='Access_Year', text_auto=True,
             category_orders={
                 'Access_Year': [f'{access}_{year}' for access in ["Closed", "Green"] for year in year_order]
             },
             color_discrete_map=color_map,
             title='期刊按访问类型和年份统计（透明度区分年份）' )

fig.update_layout(legend=dict(tracegroupgap=20))
fig.show()

2. 更优的展示方式 #

如果堆叠渐变的可读性不足，可考虑以下方案：

• 分面柱状图：按年份分面展示，每个年份单独呈现所有期刊的Access堆叠数据，年份对比更清晰：

  fig = px.bar(df, x='Journal', y='Count', color='Access', text_auto=True,
               facet_col='Year', facet_col_wrap=2,
               category_orders={'Access': ["Closed", "Green"]},
               color_discrete_map={"Closed": "#AB63FA", "Green": "#2CA02C"},
               title='期刊按访问类型统计（分年份展示）' )
  fig.show()
  

• 分组柱状图：将同一期刊的不同年份数据并列展示，而非堆叠，适合对比同一期刊跨年份的Access数据变化：

  fig = px.bar(df, x='Journal', y='Count', color='Access', text_auto=True,
               barmode='group',
               facet_col='Year', facet_col_wrap=2,
               category_orders={'Access': ["Closed", "Green"]},
               color_discrete_map={"Closed": "#AB63FA", "Green": "#2CA02C"},
               title='期刊按访问类型统计（分组展示年份）' )
  fig.show()
  

内容的提问来源于stack exchange，提问作者eschares