Thank you Salah and Phil for your answers and advice (sorry for my late
reply).
(1) Soundex works well, however I think the information provided by soundex
are not detailed enough for my purpose.
(2) The "edit distance" approach sounds promising. However during my first
tries I had some problems to fit the methodology to my dataset. I will keep
on trying and would appreciate if I could get back to you for further
questions.
Simon
-----Urspr�ngliche Nachricht-----
Von: [email protected]
[mailto:[email protected]] Im Auftrag von Phil Schumm
Gesendet: Dienstag, 3. Juni 2008 16:02
An: [email protected]
Betreff: Re: st: data problem - duplicates
On Jun 3, 2008, at 7:51 AM, <[email protected]> wrote:
> "Name" is the only variable which I can use to select duplicates. I
> know that there are ways and programs which are able to define a
> kind of "similarity-index" which holds information about how
> similar two or more variables are on the basis of counting the
> different characters between the variables.
A common way to approach this is with the concept of "edit distance,"
which is the minimum number of operations required to transform one
string into another (also known as the Levenshtein distance). I've
never implemented this in Stata myself, but a program was posted to
Statalist several years ago:
http://www.stata.com/statalist/archive/2002-08/msg00436.html
-- Phil
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
*
* For searches and help try:
* http://www.stata.com/support/faqs/res/findit.html
* http://www.stata.com/support/statalist/faq
* http://www.ats.ucla.edu/stat/stata/
