Python - Unique Values in Matrix

A matrix in Python is typically represented as a list of lists, where each inner list corresponds to a row. Finding unique values in a matrix is a common task in data processing and analysis. Python provides several efficient methods to extract unique elements from matrices.

Method 1: Using set()

The simplest approach is to iterate through all elements and add them to a set, which automatically removes duplicates ?

def unique_values_of_matrix(matrix):
    unique_elements = set()
    for row in matrix:
        for element in row:
            unique_elements.add(element)
    return unique_elements

# Example matrix
names = [
    ['John', 'Sam', 'Daniel'],
    ['Jack', 'Matt', 'John'],
    ['Sam', 'Tyler', 'Matt']
]

unique_values = unique_values_of_matrix(names)
print(sorted(unique_values))

['Daniel', 'Jack', 'John', 'Matt', 'Sam', 'Tyler']

Method 2: Using NumPy

NumPy provides the unique() function which efficiently finds unique values and returns them in sorted order ?

import numpy as np

names = [
    ['John', 'Sam', 'Daniel'],
    ['Jack', 'Matt', 'John'],
    ['Sam', 'Tyler', 'Matt']
]

# Convert to NumPy array and find unique values
matrix_array = np.array(names)
unique_values = np.unique(matrix_array)
print(unique_values)

['Daniel' 'Jack' 'John' 'Matt' 'Sam' 'Tyler']

Method 3: Using List Comprehension with Set

A more Pythonic approach using list comprehension to flatten the matrix and find unique values ?

names = [
    ['John', 'Sam', 'Daniel'],
    ['Jack', 'Matt', 'John'],
    ['Sam', 'Tyler', 'Matt']
]

# Flatten matrix and get unique values using set comprehension
unique_values = list({element for row in names for element in row})
print(sorted(unique_values))

['Daniel', 'Jack', 'John', 'Matt', 'Sam', 'Tyler']

Method 4: Using itertools.chain()

The chain() function flattens the matrix efficiently before extracting unique values ?

from itertools import chain

names = [
    ['John', 'Sam', 'Daniel'],
    ['Jack', 'Matt', 'John'],
    ['Sam', 'Tyler', 'Matt']
]

# Flatten using chain and find unique values
unique_values = set(chain(*names))
print(sorted(unique_values))

['Daniel', 'Jack', 'John', 'Matt', 'Sam', 'Tyler']

Method 5: Using Pandas

For large datasets, Pandas provides efficient methods to handle matrix operations ?

import pandas as pd

names = [
    ['John', 'Sam', 'Daniel'],
    ['Jack', 'Matt', 'John'],
    ['Sam', 'Tyler', 'Matt']
]

# Convert to DataFrame and find unique values
df = pd.DataFrame(names)
unique_values = df.stack().unique()
print(sorted(unique_values))

['Daniel', 'Jack', 'John', 'Matt', 'Sam', 'Tyler']

Performance Comparison

Conclusion

Use NumPy for numerical matrices and maximum performance. For simple cases, set() with nested loops provides clarity. Choose itertools.chain() for memory-efficient processing of large matrices.