Efficient ways to identify similar string variables

[Devin Parker]

Hello,
I am trying to identify and consolidate addresses that are technically the same, but have various string definitions.

For example:
'0105 Edwards Village Center A203'
'0105 Edwards Village Blvd.'
'0105 Edwards Village Blvd Suite A203'
'0105 Edwards Village Ctr A-203'
'0105 Edwards Village Ctr Ste A-203'

These addresses are the same but different in 'string'.
I need to collapse the addresses to '0105 Edwards Village Blvd' etc.

Is there a way to identify variables that share multiple string attributes?

Thank you.

[Andrew Musau]

There is no perfect solution for this kind of thing, but fuzzy matching can help. Search the forum for examples and implementations of this.

[Hemanshu Kumar]

The set of tools you will find useful will depend on the precise types of variations you have to consolidate. Fuzzy matching, as suggested in #2, is one way to go. From the example you have shown, I think you will also get quite far by using some basic cleaning techniques:

• convert strings to lower() case, and trim()and itrim() them to remove unnecessary spaces
• use subinstr() to remove things like hyphens and periods
• standardize acronyms: loop over subinstr() statements to standardise Ste/Suite, Ctr/Center, and so on

[Devin Parker]

Excellent! Thank you both for your suggestions.