Supersmithy
Board Regular
- Joined
- Feb 5, 2012
- Messages
- 81
Hi, I’m working with two address databases where Address1 sheet has the following headings:
Property number – Column G for the sake of below example formula
Additional Text
Add-Text2
Name
Name2
Suffix
Street – Column I
Town – Column K
Postcode
And the other, Address2 sheet has:
Property number – B2 for the sake of below example formula
Suffix
Street – D2
Name
Additional text
Town – G2
Postcode
The data is a real mess with not all fields making a logical match. Up to now I’ve been using the following multiple INDEX MATCH array formula with a minimum of 3 conditions so as to get the best match:
{=IFERROR(INDEX(All_address,MATCH(1,(Addresses1!I$1:I$226287=D2)*( Addresses1!K$1:K$226287=G2)*( Addresses1!G$1:G$226287=B2),0),1),"No Match")}
I’ve now done in the region of 30x combinations of the above formula, mostly with postcode and/or town as the basis followed by a mix of looking at name, number, additional text fields and so on.
Has anyone got any suggestions as to;
a) how I could have perhaps done this a little more efficiently? and;
b) How I could use wild cards or similar to pick up those that are a close match to as many fields as possible but not perfect? For example, some might match on postcode and number but street might have been spelt or typed out slightly differently.
I’ve got a about 20,000 lines left to match so I’m not quite prepared to resort to manual manipulation just yet!
Many Thanks
Property number – Column G for the sake of below example formula
Additional Text
Add-Text2
Name
Name2
Suffix
Street – Column I
Town – Column K
Postcode
And the other, Address2 sheet has:
Property number – B2 for the sake of below example formula
Suffix
Street – D2
Name
Additional text
Town – G2
Postcode
The data is a real mess with not all fields making a logical match. Up to now I’ve been using the following multiple INDEX MATCH array formula with a minimum of 3 conditions so as to get the best match:
{=IFERROR(INDEX(All_address,MATCH(1,(Addresses1!I$1:I$226287=D2)*( Addresses1!K$1:K$226287=G2)*( Addresses1!G$1:G$226287=B2),0),1),"No Match")}
I’ve now done in the region of 30x combinations of the above formula, mostly with postcode and/or town as the basis followed by a mix of looking at name, number, additional text fields and so on.
Has anyone got any suggestions as to;
a) how I could have perhaps done this a little more efficiently? and;
b) How I could use wild cards or similar to pick up those that are a close match to as many fields as possible but not perfect? For example, some might match on postcode and number but street might have been spelt or typed out slightly differently.
I’ve got a about 20,000 lines left to match so I’m not quite prepared to resort to manual manipulation just yet!
Many Thanks