Add ONLY the total values on top of stacked bars

Question:

I am working with the following bar plot:

Fig 1: bar plot

And I would like to add only the total amount of each index on top of the bars, like this:

Fig 2: with total amount

However, when I use the following code, I only get parts of the stacks of each bar.

import matplotlib.pyplot as plt

data = [['0.01 - 0.1','A'],['0.1 - 0.5','B'],['0.5 - 1.0','B'],['0.01 - 0.1','C'],['> 2.5','A'],['1.0 - 2.5','A'],['> 2.5','A']]

df = pd.DataFrame(data, columns = ['Size','Index'])

### plot

df_new = df.sort_values(['Index'])

list_of_colors_element = ['green','blue','yellow','red','purple']

# Draw
piv = df_new.assign(dummy=1) 
            .pivot_table('dummy', 'Index', 'Size', aggfunc='count', fill_value=0) 
            .rename_axis(columns=None)
ax = piv.plot.bar(stacked=True, color=list_of_colors_element, rot=0, width=1)

ax.bar_label(ax.containers[0],fontsize=9)

# Decorations
plt.title("Index coloured by size", fontsize=22)
plt.ylabel('Amount')
plt.xlabel('Index')
plt.grid(color='black', linestyle='--', linewidth=0.4)
plt.xticks(range(3),fontsize=15)
plt.yticks(fontsize=15)

plt.show()

I have tried with different varieties of ax.bar_label(ax.containers[0],fontsize=9) but none displays the total of the bars.

Asked By: Fred

||
Source

Answers:

As Trenton points out, bar_label is usable only if the topmost segment is never zero (i.e., exists in every stack) but otherwise not. Here are examples of the two cases.

If the topmost segment is never zero, use bar_label

In this example, the topmost segment (purple '>2.5') exists for all A, B, and C, so we can just use ax.bar_label(ax.containers[-1]):

df = pd.DataFrame({'Index': [*'AAAABBCBC'], 'Size': ['0.01-0.1', '>2.5', '1.0-2.5', '>2.5', '0.1-0.5', '0.5-1.0', '0.01-0.1', '>2.5', '>2.5']})
piv = pd.crosstab(df['Index'], df['Size'])
ax = piv.plot.bar(stacked=True)

# auto label since none of the topmost segments are missing
ax.bar_label(ax.containers[-1])

Otherwise, sum and label manually

In OP’s example, the topmost segment (purple '>2.5') does not always exist (missing for B and C), so the totals need to be summed manually.

How to compute the totals will depend on your specific dataframe. In OP’s case, A, B, and C are rows, so the totals should be computed as sum(axis=1):

df = pd.DataFrame({'Index': [*'AAAABBC'], 'Size': ['0.01-0.1', '>2.5', '1.0-2.5', '>2.5', '0.1-0.5', '0.5-1.0', '0.01-0.1']})
piv = pd.crosstab(df['Index'], df['Size'])
ax = piv.plot.bar(stacked=True)

# manually sum and label since some topmost segments are missing
for x, y in enumerate(piv.sum(axis=1)):
    ax.annotate(y, (x, y+0.1), ha='center')

Answered By: tdy