Altair分面绘图时填充颜色异常问题求助

我是Altair新手，目标是绘制带有填充点和回归线的图表，但使用facet分面功能时出现填充颜色异常的情况，推测是语法使用不够规范。

chart = alt.Chart(exercise_df).mark_point(fillOpacity=1).encode(
    x = alt.X('IntensityOfRemindersDelta'),
    y = alt.Y('HealthyEatingScoreDelta'),
    color = alt.Color('Individual'),
    fill = alt.Fill('Individual')
).properties(
    height=500,
    width=500,
)

line = (chart
    .mark_line()
    .encode(
        color=alt.value("#dddddd")
    )
    .transform_regression('IntensityOfRemindersDelta', 'HealthyEatingScoreDelta')
)

(chart + line).facet('Individual')

每个分面对应单个个体，但点的填充色并非该个体的专属颜色，而是混有其他个体的颜色，填充色显示混乱。

IndNo,Individual,IntensityOfReminders,HealthyEatingScore,IntensityOfReminder,IntensityOfRemindersMean,HealthyEatingScoreMean,IntensityOfRemindersDelta,HealthyEatingScoreDelta
1,You,1.98,0.14,0.23,3.87,0.81,-1.89,-0.67
1,You,4.58,1.37,0.54,3.87,0.81,0.71,0.56
1,You,2.81,0.77,0.33,3.87,0.81,-1.06,-0.05
1,You,2.96,0.85,0.35,3.87,0.81,-0.92,0.03
1,You,5.07,0.50,0.59,3.87,0.81,1.20,-0.31
1,You,3.14,1.27,0.37,3.87,0.81,-0.73,0.45
1,You,5.80,0.93,0.68,3.87,0.81,1.93,0.12
1,You,4.63,0.69,0.54,3.87,0.81,0.76,-0.12
2,Me,6.10,1.03,0.71,4.01,0.44,2.09,0.60
2,Me,5.06,0.21,0.59,4.01,0.44,1.05,-0.23
2,Me,2.60,0.00,0.30,4.01,0.44,-1.41,-0.44
2,Me,2.71,0.11,0.32,4.01,0.44,-1.30,-0.33
2,Me,2.13,0.57,0.25,4.01,0.44,-1.88,0.13
2,Me,2.37,0.21,0.28,4.01,0.44,-1.65,-0.23
2,Me,6.91,1.24,0.81,4.01,0.44,2.90,0.80
2,Me,4.21,0.13,0.49,4.01,0.44,0.20,-0.30
3,Shamma,5.95,0.75,0.69,6.38,0.61,-0.44,0.14
3,Shamma,4.68,0.98,0.55,6.38,0.61,-1.71,0.38
3,Shamma,7.40,0.72,0.86,6.38,0.61,1.02,0.11
3,Shamma,4.06,0.48,0.48,6.38,0.61,-2.32,-0.13
3,Shamma,7.85,0.30,0.92,6.38,0.61,1.47,-0.31
3,Shamma,7.55,0.85,0.88,6.38,0.61,1.17,0.25
3,Shamma,6.99,0.17,0.82,6.38,0.61,0.61,-0.44
3,Shamma,6.59,0.59,0.77,6.38,0.61,0.21,-0.01
4,Liqing,8.56,0.67,1.00,6.79,0.41,1.77,0.26
4,Liqing,6.89,0.61,0.81,6.79,0.41,0.10,0.19
4,Liqing,6.12,0.81,0.72,6.79,0.41,-0.67,0.39
4,Liqing,7.95,0.38,0.93,6.79,0.41,1.17,-0.04
4,Liqing,6.88,0.04,0.80,6.79,0.41,0.10,-0.37
4,Liqing,4.73,0.43,0.55,6.79,0.41,-2.05,0.02
4,Liqing,8.30,0.36,0.97,6.79,0.41,1.52,-0.05
4,Liqing,4.85,0.00,0.57,6.79,0.41,-1.93,-0.41

问题核心是分面时机与编码继承冲突，原代码先合并图表再分面，导致每个分面仍保留全局颜色映射，以下两种方式可解决：

方式1：先分面再叠加回归线 #

# 创建基础点图，暂不绑定个体颜色
base = alt.Chart(exercise_df).mark_point(fillOpacity=1).encode(
    x=alt.X('IntensityOfRemindersDelta'),
    y=alt.Y('HealthyEatingScoreDelta')
).properties(
    height=250,
    width=250
)

# 生成回归线
reg_line = base.transform_regression(
    'IntensityOfRemindersDelta', 'HealthyEatingScoreDelta'
).mark_line(color='#dddddd')

# 分面时绑定个体颜色，确保子图只显示当前个体颜色
facet_chart = base.encode(
    color=alt.Color('Individual'),
    fill=alt.Fill('Individual')
).facet(
    facet='Individual',
    columns=2  # 可选，控制分面列数
)

# 叠加回归线到每个分面子图
final_chart = facet_chart + reg_line
final_chart

方式2：用resolve_scale独立颜色刻度 #

chart = alt.Chart(exercise_df).mark_point(fillOpacity=1).encode(
    x=alt.X('IntensityOfRemindersDelta'),
    y=alt.Y('HealthyEatingScoreDelta'),
    color=alt.Color('Individual'),
    fill=alt.Fill('Individual')
).properties(
    height=250,
    width=250
)

line = chart.transform_regression(
    'IntensityOfRemindersDelta', 'HealthyEatingScoreDelta'
).mark_line(color='#dddddd')

# 分面后设置颜色/填充刻度独立，强制每个子图只显示当前个体颜色
(chart + line).facet('Individual', columns=2).resolve_scale(color='independent', fill='independent')

原因说明 #

原代码先合并图表再分面，每个分面内的点仍保留全局的Individual颜色映射规则，因此会显示所有个体的颜色。调整分面顺序或设置独立刻度，能确保每个子图仅使用对应个体的颜色编码。

内容的提问来源于stack exchange，提问作者spdrnl