I have a below table in a DataFrame. I want to visualize it in a stacked bar plot.
| Percentage | A | B | C | D | E | F |
|---|---|---|---|---|---|---|
| 0% | 10 | 20 | 50 | 60 | 100 | 90 |
| 10% | 89 | 87 | 65 | 87 | 89 | 88 |
| 50% | 87 | 876 | 89 | 978 | 6 | 0 |
| 100% | 89 | 78 | 89 | 89 | 54 | 34 |
For the above table, I want to get the stacked bar chart with x-axis as columns A,B,C,D,E,F and y-axis as Percentage Counts.
For example, x-axis column A has stacked bar with 10, 89, 87, 89 for the percentages.
I want output to be like 
Since you want dummy bars, you have to craft a dummy DataFrame:
tmp = df.set_index('Percentage').T X = np.repeat(100/tmp.shape[1], tmp.shape[1]) ax = tmp.notna().mul(X, axis=1).plot.bar(stacked=True) ax.legend(bbox_to_anchor=(1.0, 1.0)) ax.set_yticks(np.cumsum(X)-X[0]/2, tmp.columns) for k, c in zip(tmp, ax.containers): ax.bar_label(c, labels=tmp[k], label_type='center') Output:
[
Set percentage as index (set_index), transpose, and plot the stacked bars (plot.bar with stacked=True):
df.set_index('Percentage').T.plot.bar(stacked=True) Output:
If you want to group by percentage, skip the transpose and use barh:
df.set_index('Percentage').plot.barh(stacked=True) Output:
