Using Altair to generate an unstacked barplot with an already stacked data-CodePudding

I created a DataFrame with stacked data, i.e., tested \in total.

import pandas as pd
import altair as alt

df = pd.DataFrame({
  'date': ['2021-01-01', '2021-02-01', '2021-03-01'],
  'total': [10, 15, 20],
  'tested': [0, 5, 10]
})

dfm = df.melt(id_vars=['date'])

I would like to plot a stacked bar plot with Altair. Since the tested column is already contained in the total column, I would expect a chart with the max values of the total, but the result shows the sum.

alt.Chart(dfm).mark_bar().encode(
    x='date:O',
    y='value:Q',
    color='variable:O'
)

I know I can use pandas to create an untested column and generate the plot using tested and untested columns, but I would like to know if I can achieve this result without transforming the data.

CodePudding user response：

To create an unstacked bar chart you can set stack=False:

alt.Chart(dfm).mark_bar().encode(
    x='date:O',
    y=alt.Y('value:Q', stack=False),
    color='variable:O'
)

Note that it will always show the rightmost column on top (tested in the image above).