Grouped and Stacked¶

Create multi-series bar charts to compare categories across groups or show composition of totals.

Grouped Bar Chart¶

Place bars side by side to compare multiple series.

1. Basic Grouped Bars¶

```python import matplotlib.pyplot as plt import numpy as np

categories = ['A', 'B', 'C', 'D', 'E'] series1 = [23, 45, 56, 78, 32] series2 = [28, 40, 62, 70, 38]

x = np.arange(len(categories)) width = 0.35

fig, ax = plt.subplots() ax.bar(x - width/2, series1, width, label='Series 1') ax.bar(x + width/2, series2, width, label='Series 2')

ax.set_xticks(x) ax.set_xticklabels(categories) ax.legend() plt.show() ```

2. Three Groups¶

```python series1 = [23, 45, 56, 78, 32] series2 = [28, 40, 62, 70, 38] series3 = [20, 35, 50, 65, 30]

x = np.arange(len(categories)) width = 0.25

fig, ax = plt.subplots(figsize=(10, 6)) ax.bar(x - width, series1, width, label='2022') ax.bar(x, series2, width, label='2023') ax.bar(x + width, series3, width, label='2024')

ax.set_xticks(x) ax.set_xticklabels(categories) ax.legend() plt.show() ```

3. Dynamic Width Calculation¶

```python def grouped_bar(ax, data, labels, group_labels): n_groups = len(data) n_categories = len(data[0])

total_width = 0.8
bar_width = total_width / n_groups
x = np.arange(n_categories)

for i, (series, label) in enumerate(zip(data, labels)):
    offset = (i - n_groups/2 + 0.5) * bar_width
    ax.bar(x + offset, series, bar_width, label=label)

ax.set_xticks(x)
ax.set_xticklabels(group_labels)
ax.legend()

fig, ax = plt.subplots() data = [series1, series2, series3] grouped_bar(ax, data, ['2022', '2023', '2024'], categories) plt.show() ```

Pandas Grouped Bars¶

Use DataFrame's built-in plotting for automatic grouped bars.

1. DataFrame with Multiple Columns¶

```python import pandas as pd

data = {'Student': ['Brandon', 'Vanessa', 'Daniel', 'Kevin', 'William'], 'Midterm': [85, 60, 60, 65, 100], 'Final': [90, 90, 65, 80, 95]} df = pd.DataFrame(data).set_index('Student')

fig, ax = plt.subplots(figsize=(12, 3)) df.plot(kind='bar', ax=ax) ax.spines['right'].set_visible(False) ax.spines['top'].set_visible(False) plt.show() ```

2. Matplotlib Equivalent¶

```python position = np.arange(df.shape[0]) student = df.index midterm = df.Midterm final = df.Final

fig, ax = plt.subplots(figsize=(12, 3)) width = 0.3 ax.bar(position - width/2, midterm, width=width, label='Midterm') ax.bar(position + width/2, final, width=width, label='Final') ax.set_xticks(position, student) ax.set_xlabel('Student') ax.legend() ax.spines['right'].set_visible(False) ax.spines['top'].set_visible(False) plt.show() ```

3. Horizontal Grouped with Matplotlib¶

python fig, ax = plt.subplots(figsize=(12, 3)) height = 0.3 ax.barh(position - height/2, midterm, height=height, label='Midterm') ax.barh(position + height/2, final, height=height, label='Final') ax.set_yticks(position, student) ax.set_ylabel('Student') ax.legend() ax.spines['right'].set_visible(False) ax.spines['top'].set_visible(False) plt.show()

Stacked Bar Chart¶

Stack bars vertically to show composition.

1. Basic Stacked Bars¶

```python categories = ['Q1', 'Q2', 'Q3', 'Q4'] product_a = [20, 25, 30, 35] product_b = [15, 20, 25, 20] product_c = [10, 15, 10, 15]

fig, ax = plt.subplots() ax.bar(categories, product_a, label='Product A') ax.bar(categories, product_b, bottom=product_a, label='Product B') ax.bar(categories, product_c, bottom=np.array(product_a) + np.array(product_b), label='Product C')

ax.legend() ax.set_ylabel('Sales') plt.show() ```

2. Using NumPy for Bottom¶

```python product_a = np.array([20, 25, 30, 35]) product_b = np.array([15, 20, 25, 20]) product_c = np.array([10, 15, 10, 15])

fig, ax = plt.subplots() ax.bar(categories, product_a, label='Product A') ax.bar(categories, product_b, bottom=product_a, label='Product B') ax.bar(categories, product_c, bottom=product_a + product_b, label='Product C')

ax.legend() plt.show() ```

3. Dynamic Stacking¶

```python def stacked_bar(ax, data, labels, categories): bottom = np.zeros(len(categories))

for series, label in zip(data, labels):
    ax.bar(categories, series, bottom=bottom, label=label)
    bottom += np.array(series)

ax.legend()

fig, ax = plt.subplots() data = [product_a, product_b, product_c] stacked_bar(ax, data, ['Product A', 'Product B', 'Product C'], categories) plt.show() ```

Horizontal Stacked Bars¶

Stack bars horizontally using barh.

1. Basic Horizontal Stack¶

```python categories = ['Team A', 'Team B', 'Team C', 'Team D'] wins = [15, 12, 18, 10] losses = [5, 8, 2, 10] draws = [2, 2, 2, 2]

fig, ax = plt.subplots() ax.barh(categories, wins, label='Wins') ax.barh(categories, losses, left=wins, label='Losses') ax.barh(categories, draws, left=np.array(wins) + np.array(losses), label='Draws')

ax.legend() ax.set_xlabel('Games') plt.show() ```

2. Diverging Stacked Bar¶

```python categories = ['Q1', 'Q2', 'Q3', 'Q4'] positive = [30, 40, 35, 45] negative = [-20, -15, -25, -10]

fig, ax = plt.subplots() ax.barh(categories, positive, label='Gains', color='green') ax.barh(categories, negative, label='Losses', color='red') ax.axvline(x=0, color='black', linewidth=0.8) ax.legend() plt.show() ```

3. Centered Diverging¶

```python survey = ['Strongly Disagree', 'Disagree', 'Neutral', 'Agree', 'Strongly Agree'] responses = [10, 20, 15, 35, 20]

colors = ['darkred', 'red', 'gray', 'green', 'darkgreen'] starts = [0, 0, 0, 0, 0]

Center on neutral¶

starts[0] = -responses[0] - responses[1] - responses[2]/2 starts[1] = -responses[1] - responses[2]/2 starts[2] = -responses[2]/2 starts[3] = responses[2]/2 starts[4] = responses[2]/2 + responses[3]

fig, ax = plt.subplots() for i, (response, start, color, label) in enumerate(zip(responses, starts, colors, survey)): ax.barh(['Survey'], response, left=start, color=color, label=label)

ax.axvline(x=0, color='black', linewidth=0.8) ax.legend(loc='lower right') plt.show() ```

Percentage Stacked Bars¶

Show proportions that sum to 100%.

1. Calculate Percentages¶

```python categories = ['2021', '2022', '2023', '2024'] cat_a = np.array([20, 25, 30, 35]) cat_b = np.array([15, 20, 25, 20]) cat_c = np.array([10, 15, 10, 15])

totals = cat_a + cat_b + cat_c pct_a = cat_a / totals * 100 pct_b = cat_b / totals * 100 pct_c = cat_c / totals * 100 ```

2. Create Percentage Stack¶

```python fig, ax = plt.subplots() ax.bar(categories, pct_a, label='Category A') ax.bar(categories, pct_b, bottom=pct_a, label='Category B') ax.bar(categories, pct_c, bottom=pct_a + pct_b, label='Category C')

ax.set_ylabel('Percentage (%)') ax.set_ylim(0, 100) ax.legend() plt.show() ```

3. Reusable Function¶

```python def percentage_stacked_bar(ax, data, labels, categories): data = np.array(data) totals = data.sum(axis=0) percentages = data / totals * 100

bottom = np.zeros(len(categories))
for pct, label in zip(percentages, labels):
    ax.bar(categories, pct, bottom=bottom, label=label)
    bottom += pct

ax.set_ylim(0, 100)
ax.set_ylabel('Percentage (%)')
ax.legend()

fig, ax = plt.subplots() percentage_stacked_bar(ax, [cat_a, cat_b, cat_c], ['A', 'B', 'C'], categories) plt.show() ```

Adding Labels to Grouped/Stacked¶

Annotate bars with values.

1. Grouped Bar Labels¶

```python x = np.arange(len(categories)) width = 0.35

fig, ax = plt.subplots() bars1 = ax.bar(x - width/2, series1, width, label='2023') bars2 = ax.bar(x + width/2, series2, width, label='2024')

ax.bar_label(bars1, padding=3) ax.bar_label(bars2, padding=3)

ax.set_xticks(x) ax.set_xticklabels(categories) ax.legend() plt.show() ```

2. Stacked Bar Labels¶

```python fig, ax = plt.subplots() bars1 = ax.bar(categories, product_a, label='Product A') bars2 = ax.bar(categories, product_b, bottom=product_a, label='Product B') bars3 = ax.bar(categories, product_c, bottom=product_a + product_b, label='Product C')

ax.bar_label(bars1, label_type='center') ax.bar_label(bars2, label_type='center') ax.bar_label(bars3, label_type='center')

ax.legend() plt.show() ```

3. Total Labels on Stack¶

```python fig, ax = plt.subplots() bars1 = ax.bar(categories, product_a, label='Product A') bars2 = ax.bar(categories, product_b, bottom=product_a, label='Product B') bars3 = ax.bar(categories, product_c, bottom=product_a + product_b, label='Product C')

totals = product_a + product_b + product_c for i, total in enumerate(totals): ax.text(i, total + 1, str(total), ha='center', fontweight='bold')

ax.legend() plt.show() ```

Styling Grouped and Stacked¶

Apply consistent styling across multi-series charts.

1. Color Schemes¶

```python colors = ['#2ecc71', '#3498db', '#9b59b6']

fig, ax = plt.subplots() ax.bar(categories, product_a, label='A', color=colors[0]) ax.bar(categories, product_b, bottom=product_a, label='B', color=colors[1]) ax.bar(categories, product_c, bottom=product_a + product_b, label='C', color=colors[2]) ax.legend() plt.show() ```

2. Edge and Hatch¶

python fig, ax = plt.subplots() ax.bar(categories, product_a, label='A', color='white', edgecolor='blue', hatch='//') ax.bar(categories, product_b, bottom=product_a, label='B', color='white', edgecolor='green', hatch='\\\\') ax.bar(categories, product_c, bottom=product_a + product_b, label='C', color='white', edgecolor='red', hatch='xx') ax.legend() plt.show()

3. Complete Example¶

```python fig, ax = plt.subplots(figsize=(10, 6))

x = np.arange(len(categories)) width = 0.25

bars1 = ax.bar(x - width, series1, width, label='2022', color='#3498db', edgecolor='navy') bars2 = ax.bar(x, series2, width, label='2023', color='#2ecc71', edgecolor='darkgreen') bars3 = ax.bar(x + width, series3, width, label='2024', color='#e74c3c', edgecolor='darkred')

ax.set_xticks(x) ax.set_xticklabels(categories) ax.legend(title='Year') ax.set_ylabel('Value') ax.set_title('Grouped Bar Chart Comparison') ax.spines['top'].set_visible(False) ax.spines['right'].set_visible(False) ax.grid(axis='y', alpha=0.3)

plt.tight_layout() plt.show() ```

Exercises¶

Exercise 1. Create a stacked bar chart showing monthly sales of three products over four months. Use months ['Jan', 'Feb', 'Mar', 'Apr'] and values: Product A [10, 15, 12, 18], Product B [8, 12, 15, 10], Product C [5, 8, 10, 12]. Add a legend and value labels inside each bar segment.

import matplotlib.pyplot as plt
import numpy as np

months = ['Jan', 'Feb', 'Mar', 'Apr']
prod_a = [10, 15, 12, 18]
prod_b = [8, 12, 15, 10]
prod_c = [5, 8, 10, 12]

fig, ax = plt.subplots(figsize=(8, 6))
b1 = ax.bar(months, prod_a, label='Product A', color='steelblue')
b2 = ax.bar(months, prod_b, bottom=prod_a, label='Product B', color='coral')
bottom_c = [a + b for a, b in zip(prod_a, prod_b)]
b3 = ax.bar(months, prod_c, bottom=bottom_c, label='Product C', color='mediumseagreen')

ax.bar_label(b1, label_type='center', fontsize=9)
ax.bar_label(b2, label_type='center', fontsize=9)
ax.bar_label(b3, label_type='center', fontsize=9)

ax.legend()
ax.set_ylabel('Sales')
ax.set_title('Stacked Bar Chart')
plt.show()

Exercise 2. Create a 100% stacked bar chart (normalized) from the same data in Exercise 1. Convert each month's values to percentages of the total, so each bar reaches 100%. Use ax.bar with computed bottom offsets and percentage labels.

import matplotlib.pyplot as plt
import numpy as np

months = ['Jan', 'Feb', 'Mar', 'Apr']
prod_a = np.array([10, 15, 12, 18])
prod_b = np.array([8, 12, 15, 10])
prod_c = np.array([5, 8, 10, 12])
totals = prod_a + prod_b + prod_c

pct_a = prod_a / totals * 100
pct_b = prod_b / totals * 100
pct_c = prod_c / totals * 100

fig, ax = plt.subplots(figsize=(8, 6))
ax.bar(months, pct_a, label='Product A', color='steelblue')
ax.bar(months, pct_b, bottom=pct_a, label='Product B', color='coral')
ax.bar(months, pct_c, bottom=pct_a + pct_b, label='Product C', color='mediumseagreen')

ax.set_ylabel('Percentage (%)')
ax.set_title('100% Stacked Bar Chart')
ax.legend()
ax.set_ylim(0, 100)
plt.show()

Exercise 3. Create a grouped horizontal bar chart using ax.barh for comparing exam scores of three students across four subjects. Students: Alice [85, 90, 78, 92], Bob [70, 88, 95, 80], Carol [92, 75, 82, 88]. Subjects: ['Math', 'Science', 'English', 'History']. Use height=0.25 for grouping.

import matplotlib.pyplot as plt
import numpy as np

subjects = ['Math', 'Science', 'English', 'History']
alice = [85, 90, 78, 92]
bob = [70, 88, 95, 80]
carol = [92, 75, 82, 88]

y = np.arange(len(subjects))
height = 0.25

fig, ax = plt.subplots(figsize=(10, 6))
ax.barh(y - height, alice, height, label='Alice', color='steelblue')
ax.barh(y, bob, height, label='Bob', color='coral')
ax.barh(y + height, carol, height, label='Carol', color='mediumseagreen')

ax.set_yticks(y)
ax.set_yticklabels(subjects)
ax.set_xlabel('Score')
ax.set_title('Exam Scores by Student')
ax.legend()
plt.tight_layout()
plt.show()