Summary  
Shows how to implement code that calculates and applies horizontal offsets for grouped bar charts by iterating through series arrays and setting x-axis tick labels accordingly.

General domain of usage  
Data visualization

Another common option is a **grouped bar chart**, where bars for each category are placed **side by side** instead of stacking them.

This is useful when you want to compare categories **across groups** (like economic sectors in different countries), rather than within one total.

## Steps to Create a Grouped Bar Chart

1. **Set a bar width** and create an array for x-axis positions using `np.arange()`;
2. Combine your category data into a 2D array;
3. Use a `for` loop to draw each group of bars with the `bar()` function, shifting their positions horizontally;
4. Customize the x-axis tick positions and labels using `plt.xticks()`.

import matplotlib.pyplot as plt
import numpy as np

# Labels and data
countries = ['USA', 'China', 'Japan']
positions = np.arange(len(countries))

primary = np.array([1.4, 4.8, 0.4])
secondary = np.array([11.3, 6.2, 0.8])
tertiary = np.array([14.2, 8.4, 3.2])

# Group the data
sectors = np.array([primary, secondary, tertiary])

# Width of each bar
width = 0.25

# Plot each group of bars
for i in range(len(sectors)):
    plt.bar(positions + width * i, sectors[i], width)

# Center the group of bars and label the ticks
plt.xticks(positions + width, countries)

plt.show()

### How `xticks()` Works

* The **first argument** shifts the tick marks to the center of each group of bars;
* The **second argument** sets the labels using the `countries` list.

This approach works for any number of categories — just adjust the `width` to make sure the bars don't overlap.

Note

import unittest
import ast
import inspect
import user_code


def _dynamic_test(test_case, condition, success_message, failure_message):
    if condition:
        test_case._testMethodName = success_message
        test_case.assertTrue(True, success_message)
    else:
        test_case._testMethodName = failure_message
        test_case.fail(failure_message)


class TestGroupedBarChart(unittest.TestCase):

    @classmethod
    def setUpClass(cls):
        source = inspect.getsource(user_code)
        cls.tree = ast.parse(source)

        cls.for_loops = [
            node for node in ast.walk(cls.tree)
            if isinstance(node, ast.For)
        ]

        cls.bar_calls = [
            node for node in ast.walk(cls.tree)
            if isinstance(node, ast.Call)
            and isinstance(node.func, ast.Attribute)
            and node.func.attr == 'bar'
        ]

    def test_correct_array_in_len(self):
        try:
            loop = self.for_loops[0]
            arg_to_len = None

            # Check if the loop uses range(len(answers))
            if (isinstance(loop.iter, ast.Call)
                and isinstance(loop.iter.func, ast.Name)
                and loop.iter.func.id == 'range'):

                inner_call = loop.iter.args[0]
                if (isinstance(inner_call, ast.Call)
                    and isinstance(inner_call.func, ast.Name)
                    and inner_call.func.id == 'len'):

                    arg_to_len = inner_call.args[0].id if isinstance(inner_call.args[0], ast.Name) else None

            _dynamic_test(
                self,
                arg_to_len == 'answers',
                "Correct array used inside len() for loop: `len(answers)`",
                f"Expected `len(answers)`, but got `len({arg_to_len})`"
            )
        except Exception:
            self.fail("Missing or incorrect `for` loop structure.")

    def test_bar_function_used(self):
        _dynamic_test(
            self,
            len(self.bar_calls) > 0,
            "plt.bar() is used to create bars",
            "Missing plt.bar() call"
        )

    def test_correct_expression_for_x_positions(self):
        try:
            call = self.bar_calls[0]
            expr = call.args[0]
            left = expr.left.id
            right = expr.right

            _dynamic_test(
                self,
                isinstance(expr, ast.BinOp) and isinstance(expr.op, ast.Add)
                and left == 'positions'
                and isinstance(right, ast.BinOp)
                and isinstance(right.op, ast.Mult)
                and right.left.id == 'width'
                and right.right.id == 'i',
                "Correct expression for x-axis positions: positions + width * i",
                "Incorrect expression used for calculating bar positions"
            )
        except Exception:
            self.fail("Missing or invalid argument structure for x-axis positioning in plt.bar().")

    def test_correct_indexing_into_answers(self):
        try:
            call = self.bar_calls[0]
            arg = call.args[1]

            _dynamic_test(
                self,
                isinstance(arg, ast.Subscript)
                and arg.value.id == 'answers'
                and arg.slice.id == 'i',
                "Correctly indexing `answers[i]` for bar heights",
                "Incorrect indexing into answers array in plt.bar()"
            )
        except Exception:
            self.fail("Invalid or missing answers[i] indexing for bar height.")

    def test_correct_width_argument(self):
        try:
            call = self.bar_calls[0]
            width_arg = call.args[2]

            _dynamic_test(
                self,
                isinstance(width_arg, ast.Name) and width_arg.id == 'width',
                "Correct use of `width` as third argument in plt.bar()",
                f"Expected third argument to be `width`, found `{getattr(width_arg, 'id', str(width_arg))}`"
            )
        except Exception:
            self.fail("Missing or incorrect third argument in plt.bar().")


if __name__ == '__main__':
    unittest.main()

test_grouped_bar_charts.py

Data is everywhere around us, and making sense of it is extremely important. Visualization helps you deal with data by finding certain patterns and insights in it. You will develop a solid foundation of data visualization using Python and its libraries, such as matplotlib and seaborn, to get as much information from data as possible in a neat and concise way.

Discover the essentials of data visualization with Matplotlib. Learn its core concepts, explore its advantages, and create your first simple plot using this fundamental plotting library.

Master how to visualize data through the most popular plot types. Learn to build line, scatter, and bar charts to clearly communicate insights from your data.

Learn to make your plots more informative and visually appealing. Add titles, legends, colors, and grids, and discover how to arrange multiple subplots effectively.

Explore statistical visualizations that help analyze data distributions and patterns. Create histograms, box plots, and pie charts to uncover deeper statistical insights.

Take your visualization skills to the next level with Seaborn. Create advanced plots like countplots, KDEs, pair plots, and heatmaps while mastering Seaborn’s elegant style and customization options.

Grouped Bar Charts

Steps to Create a Grouped Bar Chart

How `xticks()` Works

Solution

Grouped Bar Charts

Steps to Create a Grouped Bar Chart

How xticks() Works

Solution

How `xticks()` Works