Vega/Altair直方图条形未延伸至底部问题排查

问题背景 #

我尝试使用Vega-Lite（Python中的Altair库，直接用Vega也存在相同问题）基于预分箱数据绘制直方图。数据集包含以下字段：

• bin：分箱ID
• center_point：分箱中心点
• left_bound：分箱左边界
• right_bound：分箱右边界
• n：直方图计数

使用以下Python代码绘图时，条形无法延伸至底部：

chart = alt.Chart(data).mark_bar().encode(
    x = "left_bound",
    x2 = "right_bound",
    y = "n"
)

仅当指定x2参数时会出现此问题，移除x2后条形可正常触底：

chart = alt.Chart(data).mark_bar().encode(
    x = "left_bound",
    y = "n"
)

但不能移除x2，因为分箱可能为不等宽。

问题原因 #

当同时使用x和x2定义条形宽度时，Vega-Lite默认不会强制y轴从0开始，而是根据数据中的最小n值自动调整y轴范围，导致条形看起来没有触底。

解决方案 #

显式设置y轴的刻度范围，让其从0开始。修改后的Python代码如下：

chart = alt.Chart(data).mark_bar().encode(
    x="left_bound",
    x2="right_bound",
    y=alt.Y("n", scale=alt.Scale(domain=[0, None]))
)

对应的Vega-Lite JSON规范中，需为y轴编码添加scale配置：

{"$schema": "https://vega.github.io/schema/vega-lite/v5.17.0.json","config": {"view": {"continuousHeight": 300, "continuousWidth": 500}},"data": {"name": "data-303fbb5af18e896de75409a7bbd56ac0"},"datasets": {"data-303fbb5af18e896de75409a7bbd56ac0": [{"bin": 0,"center_point": -1.95,"left_bound": -2,"n": 3401,"right_bound": -1.9},{"bin": 1,"center_point": -1.85,"left_bound": -1.9,"n": 3945,"right_bound": -1.8},{"bin": 2,"center_point": -1.75,"left_bound": -1.8,"n": 4142,"right_bound": -1.7},{"bin": 3,"center_point": -1.65,"left_bound": -1.7,"n": 4843,"right_bound": -1.6},{"bin": 4,"center_point": -1.55,"left_bound": -1.6,"n": 5440,"right_bound": -1.5},{"bin": 5,"center_point": -1.45,"left_bound": -1.5,"n": 5847,"right_bound": -1.4},{"bin": 6,"center_point": -1.35,"left_bound": -1.4,"n": 6826,"right_bound": -1.3},{"bin": 7,"center_point": -1.25,"left_bound": -1.3,"n": 7764,"right_bound": -1.2},{"bin": 8,"center_point": -1.15,"left_bound": -1.2,"n": 9205,"right_bound": -1.1},{"bin": 9,"center_point": -1.05,"left_bound": -1.1,"n": 11149,"right_bound": -1},{"bin": 10,"center_point": -0.95,"left_bound": -1,"n": 13726,"right_bound": -0.9},{"bin": 11,"center_point": -0.85,"left_bound": -0.9,"n": 17569,"right_bound": -0.8},{"bin": 12,"center_point": -0.75,"left_bound": -0.8,"n": 22626,"right_bound": -0.7},{"bin": 13,"center_point": -0.65,"left_bound": -0.7,"n": 30866,"right_bound": -0.6},{"bin": 14,"center_point": -0.55,"left_bound": -0.6,"n": 45078,"right_bound": -0.5},{"bin": 15,"center_point": -0.45,"left_bound": -0.5,"n": 72752,"right_bound": -0.4},{"bin": 16,"center_point": -0.35,"left_bound": -0.4,"n": 124899,"right_bound": -0.3},{"bin": 17,"center_point": -0.25,"left_bound": -0.3,"n": 223861,"right_bound": -0.2},{"bin": 18,"center_point": -0.15,"left_bound": -0.2,"n": 416260,"right_bound": -0.1},{"bin": 19,"center_point": -0.05,"left_bound": -0.1,"n": 1125814,"right_bound": 0},{"bin": 20,"center_point": 0.05,"left_bound": 0,"n": 445494,"right_bound": 0.1},{"bin": 21,"center_point": 0.15,"left_bound": 0.1,"n": 386848,"right_bound": 0.2},{"bin": 22,"center_point": 0.25,"left_bound": 0.2,"n": 398402,"right_bound": 0.3},{"bin": 23,"center_point": 0.35,"left_bound": 0.3,"n": 440893,"right_bound": 0.4},{"bin": 24,"center_point": 0.45,"left_bound": 0.4,"n": 537249,"right_bound": 0.5},{"bin": 25,"center_point": 0.55,"left_bound": 0.5,"n": 611479,"right_bound": 0.6},{"bin": 26,"center_point": 0.65,"left_bound": 0.6,"n": 678911,"right_bound": 0.7},{"bin": 27,"center_point": 0.75,"left_bound": 0.7,"n": 764600,"right_bound": 0.8},{"bin": 28,"center_point": 0.85,"left_bound": 0.8,"n": 927396,"right_bound": 0.9},{"bin": 29,"center_point": 0.95,"left_bound": 0.9,"n": 4172231,"right_bound": 1}]},"encoding": {"x": {"field": "left_bound", "type": "quantitative"},"x2": {"field": "right_bound"},"y": {"field": "n", "type": "quantitative", "scale": {"domain": [0, null]}}},"mark": {"type": "bar"}}

内容的提问来源于stack exchange，提问作者hemagso