Dear Victor,
This is a familiar problem, I had it with country names. I had up to
20 "variations" of the same country name. Unfortunately I had to do a
lot of stuff manually.
First I made a dictionary of "clear" values I needed. In your case
this would a "clear" list of cities. Then I tabulated my "dirty"
variable and manually assigned a correspondence between each "dirty"
value and a "clean" value. Tabulation will let you to see most common
regularities. Then saved this into a separate file and merged the
"clean" variable to the original file.
I used -regexr- and -trim- families of string functions to look for
some regularities and trim blanks. See -help string_functions-
It is always a good practice to -clonevar- your variables
I hope it helps.
Vladimir
2008/3/17, Viktor Slavtchev <[email protected]>:
> Dear list,
> I want to merge two files where the common variable is a string (names
> of cities). However, there are non systematic differences in the notions.
> For example, you can find: "Berlin" in the first file but " Berlin" in
> the second. In other cases you can find "Rome" and "Roma,IT". Or "Paris,
> FR" and "Paris/FR"
> I was tot able to find any systematics in the notion. I have over 40.000
> unique observations.
> How can I search for substrings in Stata? For example, for "*Rom*", the
> largest match between "Rome" and "Roma,IT".
> I think this could help to solve some problems. Or does anybody know a
> better way to deal with such kind of 'bad' data?
> thanks
> viktor
> *
> * 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/
>
--
__________________
Volodymyr Vakhitov
[email protected]
*
* 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/
