Click here to Skip to main content
Click here to Skip to main content

Fast, memory efficient Levenshtein algorithm

, 26 Mar 2012
Rate this:
Please Sign up or sign in to vote.
A version of the Levenshtein algorithm that uses 2*Min(StrLen1,StrLen2) bytes instead of StrLen1*StrLen2 bytes.

Introduction

The Levenshtein distance is the difference between two strings. I use it in a web crawler application to compare the new and old versions of a web page. If it has changed enough, I update it in my database.

Description

The original algorithm creates a matrix, where the size is StrLen1*StrLen2. If both strings are 1000 chars long, the resulting matrix is 1M elements; if the strings are 10,000 chars, the matrix will be 100M elements. If the elements are integers, it will be 4*100M == 400MB. Ouch!

This version of the algorithm uses only 2*StrLen elements, so the latter example would give 2*10,000*4 = 80 KB. The result is that, not only does it use less memory but it's also faster because the memory allocation takes less time. When both strings are about 1K in length, the new version is more than twice as fast.

Example

The original version would create a matrix[6+1,5+1], my version creates two vectors[6+1] (the yellow elements). In both versions, the order of the strings is irrelevant, that is, it could be matrix[5+1,6+1] and two vectors[5+1].

The new algorithm

Steps

Step Description
1 Set n to be the length of s. ("GUMBO")
Set m to be the length of t. ("GAMBOL")
If n = 0, return m and exit.
If m = 0, return n and exit.
Construct two vectors, v0[m+1] and v1[m+1], containing 0..m elements.
2 Initialize v0 to 0..m.
3 Examine each character of s (i from 1 to n).
4 Examine each character of t (j from 1 to m).
5 If s[i] equals t[j], the cost is 0.
If s[i] is not equal to t[j], the cost is 1.
6 Set cell v1[j] equal to the minimum of:
a. The cell immediately above plus 1: v1[j-1] + 1.
b. The cell immediately to the left plus 1: v0[j] + 1.
c. The cell diagonally above and to the left plus the cost: v0[j-1] + cost.
7 After the iteration steps (3, 4, 5, 6) are complete, the distance is found in the cell v1[m].

This section shows how the Levenshtein distance is computed when the source string is "GUMBO" and the target string is "GAMBOL":

Steps 1 and 2 

  v0 v1        
    G U M B O
  0 1 2 3 4 5
G 1          
A 2          
M 3          
B 4          
O 5          
L 6          

Steps 3 to 6, when i = 1

  v0 v1        
    G U M B O
  0 1 2 3 4 5
G 1 0        
A 2 1        
M 3 2        
B 4 3        
O 5 4        
L 6 5        

Steps 3 to 6, when i = 2

SWAP(v0,v1): If you look in the code you will see that I don't swap the content of the vectors but I refer to them.

Set v1[0] to the column number, e.g. 2.

    v0 v1      
    G U M B O
  0 1 2 3 4 5
G 1 0 1      
A 2 1 1      
M 3 2 2      
B 4 3 3      
O 5 4 4      
L 6 5 5      

Steps 3 to 6, when i = 3

SWAP(v0,v1).

Set v1[0] to the column number, e.g. 3.

      v0 v1    
    G U M B O
  0 1 2 3 4 5
G 1 0 1 2    
A 2 1 1 2    
M 3 2 2 1    
B 4 3 3 2    
O 5 4 4 3    
L 6 5 5 4    

Steps 3 to 6, when i = 4

SWAP(v0,v1).

Set v1[0] to the column number, e.g. 4.

        v0 v1  
    G U M B O
  0 1 2 3 4 5
G 1 0 1 2 3  
A 2 1 1 2 3  
M 3 2 2 1 2  
B 4 3 3 2 1  
O 5 4 4 3 2  
L 6 5 5 4 3  

Steps 3 to 6, when i = 5

SWAP(v0,v1).

Set v1[0] to the column number, e.g. 5.

          v0 v1
    G U M B O
  0 1 2 3 4 5
G 1 0 1 2 3 4
A 2 1 1 2 3 4
M 3 2 2 1 2 3
B 4 3 3 2 1 2
O 5 4 4 3 2 1
L 6 5 5 4 3 2

Step 7

The distance is in the lower right hand corner of the matrix, v1[m] == 2. This corresponds to our intuitive realization that "GUMBO" can be transformed into "GAMBOL" by substituting "A" for "U" and adding "L" (one substitution and one insertion = two changes).

Improvements

If you are sure that your strings will never be longer than 2^16 chars, you could use ushort instead of int, if the strings are less than 2^8 chars, you could use byte. I guess, the algorithm would be even faster if we use unmanaged code, but I have not tried it.

References

History

  • 2006-03-22
    • Version 1.0.
  • 2006-03-24
    • Detailed description of the algorithm. The code has been rewritten so that it now follows the description. Smile | :)

License

This article, along with any associated source code and files, is licensed under The Code Project Open License (CPOL)

About the Author

Sten Hjelmqvist
Software Developer (Senior) Affecto Sweden AB
Sweden Sweden
I'm a computer consultant at Affecto AB in Stockholm,Sweden and I have been designing software since the early 80's.

Comments and Discussions

 
GeneralMy vote of 5 PinmemberAmir Mehrabi-Jorshari13-Nov-13 10:06 
QuestionPerfect for my needs PinmemberGreg Lindberg5-Nov-13 3:32 
AnswerRe: Perfect for my needs PinmemberSten Hjelmqvist5-Nov-13 4:01 
GeneralMy vote of 5 Pinmembermeys_online26-Aug-12 13:38 
QuestionThanks - very interesting. PinmemberSimon Hill20-Jan-12 15:14 
AnswerRe: Thanks - very interesting. PinmemberSten Hjelmqvist25-Mar-12 23:40 
QuestionDid you port this code from somewhere? :) PinmemberMikael Svenson8-Sep-11 23:05 
Looking at the lines:
 
if (Math.Max(sRow.Length, sCol.Length) > Math.Pow(2, 31))
    throw (new Exception("\nMaximum string length in Levenshtein.iLD is " + Math.Pow(2, 31) +
    ".\nYours is " + Math.Max(sRow.Length, sCol.Length) + "."));
 
it looks to me like this is ported code. Let's examine it:
 
Math.Pow(2, 31)
is the same as int.MaxValue.
 
Math.Max(sRow.Length, sCol.Length)
will return a value <= int.MaxValue as Length from a string is of type int.
 
Thus, the whole if condition can be removed as the left side will never be greater than the right side.
 
A pretty easy spot Smile | :)
AnswerRe: Did you port this code from somewhere? :) PinmemberSten Hjelmqvist9-Sep-11 0:39 
GeneralGreat stuff Pinmembered_ward_graham28-Feb-11 5:41 
QuestionOperation Performed on strings? PinmemberPiyush Patell4-Oct-10 17:11 
GeneralVery useful. PinmemberOrganicHuman24-Aug-09 10:22 
QuestionCommercial quality? PinmemberBlake Johnston21-Nov-07 18:37 
AnswerRe: Commercial quality? PinmemberSten Hjelmqvist21-Nov-07 22:23 
NewsImprovements Available [modified] Pinmembercetinsert6-Nov-07 4:49 
GeneralRe: Improvements Available PinmemberMarc Brooks28-Jul-08 12:49 
GeneralRe: Improvements Available Pinmembercetinsert7-Aug-08 0:15 
BugRe: Improvements Available [modified] Pinmemberhbrowser4-Jul-12 4:33 
Generalattractive title Pinmemberv2.022-Mar-06 5:09 
GeneralRe: attractive title Pinmemberaprenot22-Mar-06 7:27 
GeneralRe: attractive title PinmemberThanh Dao22-Mar-06 22:16 
GeneralRe: attractive title PinmemberSten Hjelmqvist26-Mar-06 23:18 

General General    News News    Suggestion Suggestion    Question Question    Bug Bug    Answer Answer    Joke Joke    Rant Rant    Admin Admin   

Use Ctrl+Left/Right to switch messages, Ctrl+Up/Down to switch threads, Ctrl+Shift+Left/Right to switch pages.

| Advertise | Privacy | Mobile
Web02 | 2.8.140709.1 | Last Updated 26 Mar 2012
Article Copyright 2006 by Sten Hjelmqvist
Everything else Copyright © CodeProject, 1999-2014
Terms of Service
Layout: fixed | fluid