Find Similar Addresses On The Same Row

[aab425]

I need to flag addresses that are similar, but not the same, between 2 columns on the same row titled "AddressLine1" and "AddressLine2." I am working with In-DB tools. 

Ex.

AddressLine1 = 555 West Name Street, City, ARIZONA 55555

&

AddresLine2 = 555 W Name St, City, AZ 55555

I thought I had a solution, but I am getting an error (see attached).

I will appreciate any help on this!

[CarliE]

Hi @aab425,

What database are you connected to? With the in-db tools, you will have to use that databases logic instead of Alteryx logic since it is processing on the database side rather than on your machine. Depending on the connection, you may need a case statement like this:

CASE WHEN

Field = x

then x

else

x

end

Let me know, and I can try and help you solve the issue you are seeing!

Thanks,

[danilang]

Hi @aab425 

Look at this article for ideas on doing fuzzy address matching in SQL Sever.  If you're using Oracle, look here for examples of how to use UTL_MATCH

Dan

[NickWaanders]

I had kind of the same problem with counting words in descriptions of functions; we had to count the number of times 'european' and simular was mentioned, but also EU. Problem is that only a contains on EU will not give you the right answer, because in the Netherlands this is a frequently used lettre combination.

Try to make a translation table and add a space in front and back. You can use and IF statement with the Replace function in a Formula tool to change the ' ' + 'W' + ' ' to 'West' and ' ''St'' ' to 'Street'. Also check if there are simular mistakes, such as E / S / etc.

If you need any help, please let me know.

[aab425]

I am connected to SQL Server.

[aab425]

Thank you @CarliE I am connected to SQL Server.