R help - Jun 2008 - Similarity matching with probabilities

Hello,
It's just a strange coincidence that someone posted just very recently a
question about matching. I know there are several match function in the base
package (such as match, pmatch, charmatch, and the gsub etc)  but I can't
seem to use them wisely to be able to get what I need.
suppose I have the following strings:
"tets"
"estt"
"rtes7"
"gstes"
"tes5t"

Is there an R procedure to determine how related each string is to the
reference string "test", for example to say that "tets" is
similar to "test"
with a probability of 0.9 or something of that sort?
-- 
View this message in context:
http://www.nabble.com/Similarity-matching-with-probabilities-tp18153826p18153826.html
Sent from the R help mailing list archive at Nabble.com.

On 27 Jun 2008, at 14:30, francogrex wrote:
>
> Hello,
> It's just a strange coincidence that someone posted just very  
> recently a
> question about matching. I know there are several match function in  
> the base
> package (such as match, pmatch, charmatch, and the gsub etc)  but I  
> can't
> seem to use them wisely to be able to get what I need.
> suppose I have the following strings:
> "tets"
> "estt"
> "rtes7"
> "gstes"
> "tes5t"
>
> Is there an R procedure to determine how related each string is to the
> reference string "test", for example to say that "tets"
is similar
> to "test"
> with a probability of 0.9 or something of that sort?
Have a look at ?agrep.
One could loop for different max.distances to get the relation.

An other way is to calculate the edit distance by Levenshtein(- 
Damerau). A starting point could be :

http://wiki.r-project.org/rwiki/doku.php?id=tips:data-strings:levenshtein

--Hans