Distance

Warnings

• breaking The `distance` library has not been updated since November 2013 (version 0.1.3). It is considered unmaintained, and may have compatibility issues or unaddressed bugs with modern Python versions (beyond Python 3.3, which it nominally supported) and newer system architectures. Other libraries like `textdistance` have explicitly removed support for 'distance' due to its unmaintained status.
  Fix: Consider using actively maintained alternatives such as `textdistance`, `python-Levenshtein` (now `Levenshtein`), `editdistance`, or `scipy.spatial.distance` for similar functionality, which offer better compatibility and performance with current Python environments.
• gotcha Installing the optional C extension requires a C compiler and Python development headers to be present on your system. Without these, the `pip install` command with `--with-c` will fail. The benefit of the C extension for performance might be negligible or detrimental on modern Python versions due to potential inefficiencies in old C API usage.
  Fix: If C extension installation fails, remove the `--global-option="--with-c"` flag and use the pure Python implementation. For better performance with C extensions, consider modern libraries that are actively maintained and provide optimized C/Cython implementations (e.g., `Levenshtein` for string distances).
• gotcha The library's changelog from 2013 mentions switching 'back to using the to-be-deprecated Python unicode api' for Python 2.7+ compatibility and fixing 'variable interversions in (C) levenshtein which produced sometimes strange results'. This indicates potential historical issues with string handling, especially Unicode, which could resurface or behave unexpectedly in certain edge cases with modern Python 3.x string types.
  Fix: Thoroughly test distance calculations with diverse string inputs, especially those involving non-ASCII characters or complex Unicode, if you choose to use this library. Prefer modern libraries for robust Unicode support.

Install

• pip install Distance Install stable version
• pip install Distance --global-option="--with-c" Install with C extension (requires C compiler and Python headers)

Imports

• levenshtein

  import distance
  distance.levenshtein('string1', 'string2')

• hamming

  import distance
  distance.hamming('string1', 'string2')

• jaccard

  import distance
  distance.jaccard('seq1', 'seq2')

• sorensen

  import distance
  distance.sorensen('seq1', 'seq2')

Quickstart

Calculates Levenshtein, Hamming, and Jaccard distances between sequences using the primary functions provided by the library.

import distance

# Levenshtein Distance
word1 = "kitten"
word2 = "sitting"
lev_dist = distance.levenshtein(word1, word2)
print(f"Levenshtein distance between '{word1}' and '{word2}': {lev_dist}")

# Hamming Distance (for sequences of equal length)
seq1 = "karolin"
seq2 = "kathrin"
ham_dist = distance.hamming(seq1, seq2)
print(f"Hamming distance between '{seq1}' and '{seq2}': {ham_dist}")

# Jaccard Distance
set1 = "apple"
set2 = "apply"
jacc_dist = distance.jaccard(set1, set2)
print(f"Jaccard distance between '{set1}' and '{set2}': {jacc_dist}")