numpy.percentile() in python

numpy.percentile()function used to compute the nth percentile of the given data (array elements) along the specified axis.

Syntax : numpy.percentile(arr, n, axis=None, out=None)

Parameters :
arr :input array.
n : percentile value.
axis : axis along which we want to calculate the percentile value. Otherwise, it will consider arr to be flattened(works on all the axis). axis = 0 means along the column and axis = 1 means working along the row.
out :Different array in which we want to place the result. The array must have same dimensions as expected output.

Return :nth Percentile of the array (a scalar value if axis is none)or array with percentile values along specified axis.

Example 1: Compute Percentiles for a 1D Array

import numpy as np

# 1D array
arr = [20, 2, 7, 1, 34]

print("Array:", arr)
print("50th Percentile (Median):", np.percentile(arr, 50))
print("25th Percentile (Q1):", np.percentile(arr, 25))
print("75th Percentile (Q3):", np.percentile(arr, 75))

Output:

 Array: [20, 2, 7, 1, 34] 50th Percentile (Median): 7.0 25th Percentile (Q1): 2.0 75th Percentile (Q3): 20.0 

The 50th percentile equals the median, while 25th and 75th represent lower and upper quartiles respectively.

Example 2: Percentiles on a 2D Array

import numpy as np

arr = [
    [14, 17, 12, 33, 44],
    [15, 6, 27, 8, 19],
    [23, 2, 54, 1, 4]
]

print("Array:\n", arr)

# Flattened percentile (all elements)
print("\n50th Percentile (Flattened):", np.percentile(arr, 50))
print("0th Percentile (Min):", np.percentile(arr, 0))

# Percentile along columns (axis=0)
print("\n50th Percentile along axis=0:", np.percentile(arr, 50, axis=0))
print("0th Percentile along axis=0:", np.percentile(arr, 0, axis=0))

Output:

 Array: [[14, 17, 12, 33, 44], [15, 6, 27, 8, 19], [23, 2, 54, 1, 4]] 50th Percentile (Flattened): 15.0 0th Percentile (Min): 1.0 50th Percentile along axis=0: [15. 6. 27. 8. 19.] 0th Percentile along axis=0: [14. 2. 12. 1. 4.] 

axis=0 calculates percentiles column-wise, treating each column independently.

Example 3: Row-Wise Percentiles (axis=1)

import numpy as np

arr = [
    [14, 17, 12, 33, 44],
    [15, 6, 27, 8, 19],
    [23, 2, 54, 1, 4]
]

# Percentile along rows
print("50th Percentile (Median) along axis=1:", np.percentile(arr, 50, axis=1))
print("0th Percentile (Min) along axis=1:", np.percentile(arr, 0, axis=1))

# Keep same dimensions
print("\nUsing keepdims=True")
print("50th Percentile with keepdims:\n", np.percentile(arr, 50, axis=1, keepdims=True))
print("0th Percentile with keepdims:\n", np.percentile(arr, 0, axis=1, keepdims=True))

Output:

 50th Percentile (Median) along axis=1: [17. 15. 4.] 0th Percentile (Min) along axis=1: [12. 6. 1.] Using keepdims=True 50th Percentile with keepdims: [[17.] [15.] [ 4.]] 0th Percentile with keepdims: [[12.] [ 6.] [ 1.]] 

Setting keepdims=True preserves the original array’s dimensionality, useful for aligning results in matrix computations.

Example 4: Multiple Percentiles at Once

import numpy as np

arr = np.array([10, 20, 30, 40, 50, 60, 70])
percentiles = [10, 25, 50, 75, 90]
print("Array:", arr)
print("Percentiles:", np.percentile(arr, percentiles))

Output:

 Array: [10 20 30 40 50 60 70] Percentiles: [16. 25. 40. 55. 64.] 

You can compute multiple percentile values in a single call by passing a list of percentile values.

Summary

numpy.percentile() computes the nth percentile of a dataset.

Works on multi-dimensional arrays and supports flattening and broadcasting.

Use axis, keepdims, and interpolation options for flexible data summarization.

Useful for statistical analysis, outlier detection, and machine learning preprocessing.