Summary  
This chapter covers how to use NumPy’s concatenate function to join multiple arrays along a specified axis, forming larger arrays.  

General domain of usage  
Machine learning

**Array concatenation** is a fundamental operation in **NumPy** that combines arrays along a specified axis to create larger, more comprehensive datasets. This is especially useful in **machine learning**, where data is often split across multiple arrays or stored separately, such as when it comes from **different sources**.

Larger, unified datasets typically improve the performance of ML models and neural networks.

Study More

Essentially, **concatenation** involves joining arrays together to form a new array.

**NumPy** has a `concatenate()` function that enables you to concatenate arrays along a specified axis:
* `axis=0` (the default value) concatenates the arrays by rows;
* `axis=1` concatenates the arrays by columns.

The first parameter of this function is the **sequence of arrays** (a `tuple` or `list` of arrays) to concatenate, while `axis` is the second parameter.

import numpy as np
array1 = np.array([1, 2, 3])
array2 = np.array([4, 5, 6])
# Concatenating 1D arrays along their only axis 0
concatenated_array = np.concatenate((array1, array2))
print(concatenated_array)

Concatenation creates a 1D array with the elements of the first array **followed** by the elements of the second array.

Concatenating 2D arrays is performed in a similar way, but you also have to specify the `axis` parameter:

import numpy as np
array1 = np.array([[1, 2], [3, 4]])
array2 = np.array([[5, 6], [7, 8]])
# Concatenating along the axis 0 (rows)
concatenated_array_rows = np.concatenate((array1, array2))
print(f'Axis = 0:\n{concatenated_array_rows}')
# Concatenating along the axis 1 (columns)
concatenated_array_columns = np.concatenate((array1, array2), axis=1)
print(f'Axis = 1:\n{concatenated_array_columns}')

The purple elements correspond to `array1`, and the green ones to `array2`.

In fact, we can concatenate any number of arrays, and it will work the same way.

import unittest
import importlib.util
import numpy as np
import io
from contextlib import redirect_stdout


def _dynamic_test(test_case, condition, success_message, failure_message):
    """Codefinity-style test output"""
    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 TestUserCode(unittest.TestCase):
    @classmethod
    def setUpClass(cls):
        # Початкові дані
        cls.sales_data_2021 = np.array([[350, 420, 380, 410],
                                        [270, 320, 290, 310]])
        cls.sales_data_2022 = np.array([[370, 430, 400, 390],
                                        [280, 330, 300, 370]])
        # Очікуваний результат об’єднання по стовпцях
        cls.expected_combined = np.concatenate(
            (cls.sales_data_2021, cls.sales_data_2022), axis=1
        )

    def test_concatenation_by_columns(self):
        """1–2. Sales data concatenated by columns with 2022 data following 2021"""
        spec = importlib.util.spec_from_file_location("user_code", "user_code.py")
        module = importlib.util.module_from_spec(spec)
        spec.loader.exec_module(module)

        ok = hasattr(module, "combined_sales_by_product")
        arr = getattr(module, "combined_sales_by_product", None)
        type_ok = isinstance(arr, np.ndarray)
        shape_ok = type_ok and arr.shape == self.expected_combined.shape
        val_ok = type_ok and np.array_equal(arr, self.expected_combined)

        # Перевірка, що об’єднання відбулося саме по стовпцях (axis=1)
        correct_axis = type_ok and arr.shape[1] == (
            self.sales_data_2021.shape[1] + self.sales_data_2022.shape[1]
        )

        # Перевірка, що друга половина — це саме дані 2022 року
        follows_ok = type_ok and np.array_equal(
            arr[:, -self.sales_data_2022.shape[1]:], self.sales_data_2022
        )

        _dynamic_test(
            self,
            ok and type_ok and shape_ok and val_ok and correct_axis and follows_ok,
            "Sales data concatenated correctly by columns with 2022 data following 2021",
            "Sales data concatenation incorrect (axis or order wrong)",
        )

    def test_print_output(self):
        """Перевірка, що результат надруковано"""
        f = io.StringIO()
        spec = importlib.util.spec_from_file_location("user_code", "user_code.py")
        module = importlib.util.module_from_spec(spec)
        with redirect_stdout(f):
            spec.loader.exec_module(module)
        output = f.getvalue()

        printed_phrase = "Combined sales by product" in output
        printed_values = all(str(x) in output for x in [350, 430, 370])
        _dynamic_test(
            self,
            printed_phrase and printed_values,
            "Combined sales printed correctly",
            "Print statement missing or incorrect",
        )


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


test_code.py

Master practical NumPy skills for data science by focusing on multidimensional array manipulation, vectorized mathematical operations, workflow optimization, and complex matrix building, with action-driven chapters and hands-on challenges.

Array Concatenation

解答