View Full Version : compare 1 out of 1,600,000

Felix Atagong
02-14-2008, 07:08 AM
I get daily Excel (2000) files containing +/- 100 addresses in the uk. I need to compare the postal code (and city) on this file with the official British Post database that is 80 MB big (in csv) and contains about 1,600,000 entries. (This to have the official spelling of the city name, BIRMINGHAM not B'HAM etc... and to see if the right postal code has been given).

I suppose it will be faster to lookup the postal codes when they are in a database such as msaccess, not csv. Are there 'faster' databases (SQLite?). I'm basically looking for the fastest way to do a lookup...

02-14-2008, 09:18 AM
If you loaded it in access, you could have your table for the day with your +/- 100 addresses then your correct city / postal code table loaded in as well.

Then do a query with the two tables including all records from your daily table and only those that match from the other table with the relationship between postal codes. Load all your fields from the daily table into the fields section of the query. In the criteria for city use something like <>[masterpostcodetable].[city] so it will only show you records where there are differences

02-14-2008, 02:26 PM
Do you HAVE to use THAT database file?

Could you perform a web query (http://www.15seconds.com/issue/991021.htm) instead?