Introduction

The median is the middle value of an ordered data set, which makes it a useful summary when outliers would distort the mean. In Python, the best way to compute it depends on whether you want the simplest built-in tool, a manual implementation, or array-oriented performance with NumPy.

The Easiest Option: statistics.median

For normal Python lists, the standard library already provides exactly what you want:

1from statistics import median
2
3values = [7, 1, 5, 9, 3]
4print(median(values))

This works for both odd and even list lengths:

1from statistics import median
2
3print(median([1, 3, 5]))      # 3
4print(median([1, 3, 5, 7]))   # 4.0

For an odd number of items, the median is the center value after sorting. For an even number of items, Python returns the average of the two middle values.

This is the most readable solution and should be your default choice unless you have a specific reason not to use the standard library.

How to Compute It Manually

If you want to understand the logic or avoid an import, the algorithm is simple:

1. sort the list
2. find the middle index
3. return the center value, or average the two center values

1def median_manual(values):
2    if not values:
3        raise ValueError("median requires at least one value")
4
5    sorted_values = sorted(values)
6    n = len(sorted_values)
7    mid = n // 2
8
9    if n % 2 == 1:
10        return sorted_values[mid]
11    return (sorted_values[mid - 1] + sorted_values[mid]) / 2
12
13
14print(median_manual([7, 1, 5, 9, 3]))
15print(median_manual([1, 3, 5, 7]))

The important detail is using sorted(values) instead of values.sort() if you do not want to mutate the original list.

Using NumPy for Array Workloads

If your data is already in a NumPy array, numpy.median is the natural option:

1import numpy as np
2
3arr = np.array([7, 1, 5, 9, 3])
4print(np.median(arr))

NumPy is especially useful when median is part of a larger numerical workflow, or when you want to compute medians across an axis in a matrix:

1import numpy as np
2
3matrix = np.array([
4    [1, 10, 3],
5    [4, 5, 6],
6    [7, 8, 9],
7])
8
9print(np.median(matrix))
10print(np.median(matrix, axis=0))
11print(np.median(matrix, axis=1))

That is much more convenient than manually slicing nested lists when you are doing data analysis or scientific computing.

Choosing the Right Approach

Use statistics.median for ordinary Python programs and scripts. Use a manual version if you are learning the algorithm or need full control over how the values are handled. Use NumPy when the data is already array-based or when you need axis-aware operations.

The mathematical definition stays the same in all three cases. The main difference is ergonomics and the data structures you are already working with.

Common Pitfalls

The biggest mistake is forgetting to sort the data before taking the middle element manually. The median is defined on ordered values, so taking the center of the original list is usually wrong.

Another common issue is handling even-length lists incorrectly. The median is not the lower middle or upper middle item by default; it is the average of the two central values.

Empty input is another edge case. statistics.median([]) and a sensible manual implementation should raise an error because there is no meaningful median for an empty data set.

Mixed non-numeric types can also cause trouble. If the list contains values that cannot be ordered or averaged together, median calculation will fail or produce results you do not want.

Summary

• 'statistics.median is the simplest way to find the median of a Python list.'
• A manual implementation works by sorting and selecting the middle value or middle pair.
• 'numpy.median is a better fit for arrays and axis-based calculations.'
• Remember that even-length lists use the average of the two middle values.
• Handle empty input explicitly instead of assuming a median always exists.