Grouped bar charts in Matplotlib - Time & Space Complexity

Time Complexity: Grouped bar charts

O(n)

Understanding Time Complexity

When we create grouped bar charts with matplotlib, we want to know how the time to draw the chart changes as we add more groups or bars.

We ask: How does the drawing time grow when the data size grows?

Scenario Under Consideration

Analyze the time complexity of the following code snippet.

import matplotlib.pyplot as plt

labels = ['A', 'B', 'C', 'D']
values1 = [5, 7, 3, 4]
values2 = [6, 8, 4, 5]

x = range(len(labels))
width = 0.35

plt.bar([i - width/2 for i in x], values1, width=width, label='Group 1')
plt.bar([i + width/2 for i in x], values2, width=width, label='Group 2')
plt.legend()
plt.show()

This code draws two groups of bars side by side for each label.

Identify Repeating Operations

Identify the loops, recursion, array traversals that repeat.

• Primary operation: Drawing bars for each group and label.
• How many times: Once per bar, so number of groups times number of labels.

How Execution Grows With Input

As the number of labels or groups grows, the number of bars to draw grows too.

Input Size (n = labels x groups)	Approx. Operations
10 (e.g., 5 labels x 2 groups)	About 10 bar draws
100 (e.g., 50 labels x 2 groups)	About 100 bar draws
1000 (e.g., 500 labels x 2 groups)	About 1000 bar draws

Pattern observation: The drawing work grows directly with the total number of bars.

Final Time Complexity

Time Complexity: O(n)

This means the time to draw grows in a straight line with the number of bars.

Common Mistake

[X] Wrong: "Adding more groups does not affect drawing time much because bars are drawn together."

[OK] Correct: Each bar is drawn separately, so more groups mean more bars and more drawing steps.

Interview Connect

Understanding how drawing time grows helps you explain performance when visualizing large datasets, a useful skill in data science roles.

Self-Check

"What if we added a third group of bars? How would the time complexity change?"

Practice

1. What is the main purpose of a grouped bar chart in data visualization?

easy

A. To compare multiple categories side by side within groups

B. To show the trend of a single variable over time

C. To display parts of a whole as percentages

D. To plot data points on a continuous scale

Solution

1. Step 1: Understand the chart type

   Grouped bar charts place bars for different categories next to each other for easy comparison.

2. Step 2: Identify the purpose

   This helps compare multiple categories side by side within groups, unlike line or pie charts.

3. Final Answer:

   To compare multiple categories side by side within groups -> Option A

4. Quick Check:

   Grouped bars = side-by-side comparison [OK]

Hint: Grouped bars show categories side by side for easy comparison [OK]

Common Mistakes:

• Confusing grouped bar charts with line charts
• Thinking grouped bars show parts of a whole
• Assuming grouped bars plot continuous data

2. Which of the following code snippets correctly shifts bars to create a grouped bar chart in matplotlib?

easy

A. plt.bar(x - width/2, data1, width); plt.bar(x + width/2, data2, width)

B. plt.bar(x, data1, width); plt.bar(x, data2, width)

C. plt.bar(x + width, data1, width); plt.bar(x + width, data2, width)

D. plt.bar(x - width, data1, width); plt.bar(x - width, data2, width)

Solution

1. Step 1: Understand bar shifting for grouping

   To group bars side by side, shift one set left by half the bar width and the other right by half the bar width.

2. Step 2: Check the code

   plt.bar(x - width/2, data1, width); plt.bar(x + width/2, data2, width) shifts bars by -width/2 and +width/2, correctly grouping them.

3. Final Answer:

   plt.bar(x - width/2, data1, width); plt.bar(x + width/2, data2, width) -> Option A

4. Quick Check:

   Shift bars ±width/2 for grouping [OK]

Hint: Shift bars by ±width/2 to group them side by side [OK]

Common Mistakes:

• Not shifting bars, causing overlap
• Shifting bars by full width, creating gaps
• Using the same x positions for all bars

3. What will be the output of the following code?

import matplotlib.pyplot as plt
import numpy as np

labels = ['A', 'B', 'C']
data1 = [5, 7, 9]
data2 = [6, 8, 10]
x = np.arange(len(labels))
width = 0.4

plt.bar(x - width/2, data1, width, label='Data1')
plt.bar(x + width/2, data2, width, label='Data2')
plt.xticks(x, labels)
plt.legend()
plt.show()

medium

A. A stacked bar chart with Data2 bars on top of Data1 bars

B. A grouped bar chart with bars for Data1 and Data2 side by side for each label

C. A line chart connecting data points for Data1 and Data2

D. A scatter plot with points at positions x and data values

Solution

1. Step 1: Analyze the bar positions

   Bars for data1 are at x - width/2, and for data2 at x + width/2, placing them side by side.

2. Step 2: Understand the plot type

   Using plt.bar twice with shifted x positions creates a grouped bar chart, not stacked or line chart.

3. Final Answer:

   A grouped bar chart with bars for Data1 and Data2 side by side for each label -> Option B

4. Quick Check:

   Shifted bars = grouped bar chart [OK]

Hint: Shift bars left and right to get grouped bars, not stacked [OK]

Common Mistakes:

• Thinking bars are stacked instead of grouped
• Confusing bar chart with line or scatter plot
• Ignoring the x-axis label positions

4. Identify the error in this grouped bar chart code:

import matplotlib.pyplot as plt
import numpy as np

labels = ['X', 'Y', 'Z']
data1 = [3, 5, 7]
data2 = [4, 6, 8]
x = np.arange(len(labels))
width = 0.3

plt.bar(x, data1, width, label='Data1')
plt.bar(x + width/2, data2, width, label='Data2')
plt.xticks(x, labels)
plt.legend()
plt.show()

medium

A. Legend labels are missing

B. Width is too large causing bars to be invisible

C. x positions are incorrect and cause an error

D. Bars overlap because data1 bars are not shifted left

Solution

1. Step 1: Check bar positions

   data1 bars are at x, data2 bars at x + width/2, so data1 bars are not shifted left to avoid overlap.

2. Step 2: Understand grouping logic

   To group bars, data1 should be at x - width/2 and data2 at x + width/2; otherwise, bars overlap.

3. Final Answer:

   Bars overlap because data1 bars are not shifted left -> Option D

4. Quick Check:

   Shift both bars ±width/2 to avoid overlap [OK]

Hint: Shift first bars left by width/2 to prevent overlap [OK]

Common Mistakes:

• Not shifting first bar set left
• Using full width shift causing gaps
• Ignoring bar width when setting positions

5. You have sales data for 3 products over 4 quarters:

products = ['P1', 'P2', 'P3']
quarters = ['Q1', 'Q2', 'Q3', 'Q4']
sales = [[10, 15, 20, 25], [12, 18, 22, 28], [14, 20, 24, 30]]

How would you create a grouped bar chart showing sales per quarter for each product using matplotlib?

hard

A. Use a line plot instead of bars for grouped comparison

B. Plot bars at x positions equal to product indices without shifting

C. Use x = np.arange(len(quarters)), width = 0.25, then plot bars for each product shifted by -width, 0, +width

D. Stack bars for all products at the same x positions for quarters

Solution

1. Step 1: Set x positions for quarters

   Use x = np.arange(len(quarters)) to represent each quarter on the x-axis.

2. Step 2: Define bar width and shifts

   Set width = 0.25 and shift bars for each product by -width, 0, and +width to group them side by side.

3. Step 3: Plot bars for each product

   Plot bars for each product's sales data at shifted x positions to create grouped bars per quarter.

4. Final Answer:

   Use x = np.arange(len(quarters)), width = 0.25, then plot bars for each product shifted by -width, 0, +width -> Option C

5. Quick Check:

   Shift bars by ±width around x for grouping [OK]

Hint: Shift bars by ±width around x for each product group [OK]

Common Mistakes:

• Not shifting bars causing overlap
• Using product indices as x instead of quarters
• Stacking bars instead of grouping