Altair- 显示所有轴刻度但仅显示一些刻度标签

Question

I'm stumped on how to display ticks for all x-axis values but only labels for some. The x-axis values in my chart are 0.0 - 3.0 in tenths, whereas I only want to display labels for the whole numbers (0, 1, 2, 3).

Here's the data I'm working with:

bin count
0.0 0
0.1 0
0.2 0
0.3 0
0.4 0
0.5 0
0.6 0
0.7 0
0.8 0
0.9 0
1.0 0
1.1 1
1.2 3
1.3 7
1.4 14
1.5 13
1.6 29
1.7 47
1.8 59
1.9 59
2.0 75
2.1 72
2.2 103
2.3 96
2.4 119
2.5 76
2.6 93
2.7 68
2.8 70
2.9 44
3.0 49

The only progress I could manage was setting the data type to Ordinal and increasing the label font size:

alt.Chart(df).mark_bar().encode(x=alt.X('bin:O', axis=alt.Axis(labelFontSize=18, 
                                                  values=[round(i,1) for i in np.arange(0,3.1,.1)])), 
                                y=alt.Y('count:Q'))

This is what the chart looks like: altair bar chart

Any suggestions?

Answer 1

You can use the tickCount axis property to specify how many ticks you would like, and the labels are automatically chosen so that they will not overlap:

import altair as alt
import pandas as pd

df = pd.DataFrame({
    'bin': [0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0, 1.1, 1.2, 1.3, 1.4, 1.5,
            1.6, 1.7, 1.8, 1.9, 2.0, 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7, 2.8, 2.9, 3.0],
    'count': [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 3, 7, 14, 13, 29, 47,
              59, 59, 75, 72, 103, 96, 119, 76, 93, 68, 70, 44, 49]
})

alt.Chart(df).mark_bar().encode(
    x=alt.X('bin:Q', axis=alt.Axis(tickCount=df.shape[0], grid=False)),
    y=alt.Y('count:Q')
)

If you want to further customize the label locations and text, you can use a labelExpr . For example:

alt.Chart(df).mark_bar().encode(
    x=alt.X('bin:Q', axis=alt.Axis(
        tickCount=df.shape[0],
        grid=False,
        labelExpr="datum.value % 1 ? null : datum.label"
    )),
    y=alt.Y('count:Q')
)