Comments on: Damerau-Levenshtein Distance in Python http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/ Keeping track of what I do Fri, 30 Jul 2010 09:05:49 +0000 http://wordpress.org/?v=2.9.1 hourly 1 By: numerix http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-34980 numerix Thu, 27 May 2010 13:11:11 +0000 http://www.guyrutenberg.com/?p=182#comment-34980 Thanks Michael, for pointing that out. I was not aware of that difference! Thanks Michael, for pointing that out. I was not aware of that difference!

]]> By: Michael http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-34876 Michael Tue, 25 May 2010 02:54:14 +0000 http://www.guyrutenberg.com/?p=182#comment-34876 Oh, and to Guy, yes, there's a problem in the algorithm - try it on "AB" and "BA". It should give one but instead says two. I didn't figure out why that was. Oh, and to Guy, yes, there’s a problem in the algorithm – try it on “AB” and “BA”. It should give one but instead says two. I didn’t figure out why that was.

]]> By: Michael http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-34875 Michael Tue, 25 May 2010 02:49:18 +0000 http://www.guyrutenberg.com/?p=182#comment-34875 numerix, note the difference between Levenshtein Distance and Damerau-Levenshtein distance. The latter allows transpositions. The former is what is described in the SPOJ problem, and in the calculator you link to. Both algorithms give the correct answer for that pair: delete H, U, R, replace B with K, transpose O and H, replace P with Z. Six steps. numerix, note the difference between Levenshtein Distance and Damerau-Levenshtein distance. The latter allows transpositions. The former is what is described in the SPOJ problem, and in the calculator you link to.

Both algorithms give the correct answer for that pair: delete H, U, R, replace B with K, transpose O and H, replace P with Z. Six steps.

]]> By: numerix http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-34839 numerix Mon, 24 May 2010 11:53:10 +0000 http://www.guyrutenberg.com/?p=182#comment-34839 Yes, the algorithm shown above is not correct, but the other one J. Reagle gave a link to, is also incorrect. Example: The Levenshtein-Distance between HURQBOHP and QKHOZ is 7, but both algorithms give 6 as result. You can check it e.g. here: http://www-igm.univ-mlv.fr/~lecroq/seqcomp/node2.html It is true, that the second algorithm is (by far!) faster than the one shown above (I checked it with two strings of length 2000 and it was about 100 times faster), but both algorithms are unefficient implementations. My own Python implementation is about 1000 times faster than the one above. I won't show the code, because I used it for SPOJ problem EDIST (https://www.spoj.pl/problems/EDIST/). You can use that problem to check correctness and performance of Levenshtein-algorithms. Yes, the algorithm shown above is not correct, but the other one J. Reagle gave a link to, is also incorrect.
Example: The Levenshtein-Distance between HURQBOHP and QKHOZ is 7, but both algorithms give 6 as result.
You can check it e.g. here: http://www-igm.univ-mlv.fr/~lecroq/seqcomp/node2.html

It is true, that the second algorithm is (by far!) faster than the one shown above (I checked it with two strings of length 2000 and it was about 100 times faster), but both algorithms are unefficient implementations. My own Python implementation is about 1000 times faster than the one above. I won’t show the code, because I used it for SPOJ problem EDIST (https://www.spoj.pl/problems/EDIST/). You can use that problem to check correctness and performance of Levenshtein-algorithms.

]]> By: Joseph Reagle http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-23991 Joseph Reagle Fri, 09 Oct 2009 14:47:03 +0000 http://www.guyrutenberg.com/?p=182#comment-23991 Thanks for this, however I found another which is faster. http://mwh.geek.nz/2009/04/26/python-damerau-levenshtein-distance/ Thanks for this, however I found another which is faster. http://mwh.geek.nz/2009/04/26/python-damerau-levenshtein-distance/

]]> By: Guy http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-17163 Guy Tue, 17 Mar 2009 12:26:55 +0000 http://www.guyrutenberg.com/?p=182#comment-17163 @Sloan, I've just check the English wikipedia article about Damerau-Levenshtein distance (and the talk page) and didn't see anyone referring to a mistake in my implementation. Can you please point out the error, if it really exists, so I can correct it? @Sloan, I’ve just check the English wikipedia article about Damerau-Levenshtein distance (and the talk page) and didn’t see anyone referring to a mistake in my implementation.

Can you please point out the error, if it really exists, so I can correct it?

]]> By: Sloan http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-17162 Sloan Tue, 17 Mar 2009 12:18:27 +0000 http://www.guyrutenberg.com/?p=182#comment-17162 incorrect algorithm refer to wikipedia they say that whatever you have written is worng incorrect algorithm
refer to wikipedia
they say that whatever you have written is worng

]]> By: Guy http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-6192 Guy Wed, 17 Dec 2008 11:06:03 +0000 http://www.guyrutenberg.com/?p=182#comment-6192 אני משתמש בתוסף בשם wp-syntax הוא תומך בהרבה שפות תכנות והוא נוח לשימוש אני משתמש בתוסף בשם
wp-syntax
הוא תומך בהרבה שפות תכנות והוא נוח לשימוש

]]> By: אריאל http://www.guyrutenberg.com/2008/12/15/damerau-levenshtein-distance-in-python/comment-page-1/#comment-6191 אריאל Wed, 17 Dec 2008 10:52:00 +0000 http://www.guyrutenberg.com/?p=182#comment-6191 אני אהיה אוף-טופיק לגמרי, כיצד עיצבת את מראה הקוד הנ"ל, האם אתה משתמש בתוסף כלשהו שעושה את העבודה? אני אהיה אוף-טופיק לגמרי, כיצד עיצבת את מראה הקוד הנ”ל, האם אתה משתמש בתוסף כלשהו שעושה את העבודה?

]]>