Dealing with "almost" duplicate records

  • Thread starter Thread starter Guest
  • Start date Start date
G

Guest

One of my tables (originally designed in Excel) has about 17 fields. I have
several hundred (or even several thousand) records that differ by one field
entry. For me, the best way would be to create an 18th field so that whenever
two records are nearly duplicate in this manner, I can append the "different"
value to that 18th field. Otherwise, that 18th field would have a null value
for all other records. Any ideas on how to accomplish this?

I should also mention that after appending the "different" value to an 18th
field, I want to delete the record that I appended from. Make sense?
 
How do you propose to define "nearly duplicate"? If we were discussing a
"person" table, I'd ask if you considered the following to be "nearly
duplicate":

Chas. Jones
Charles A. Jones
Cahrles Jone
Charlie Jones
Chuck Jones
Charles A. Jones, III

If you want to use automated tools (Access & Excel functions), you'll need
to have it tightly defined.

Or you could add a Yes/No field and approach this with USB (using someone's
brain) -- inspect every record and check those you'll need to process
further.
 
The records are nearly duplicate in the sense 16 of the 17 fields match. The
17th field is indeed different with a unique meaning that needs to preserved.
Thus, I want to make an 18th field to hold that extra value from the 17th
field, then delete the extra record.
 
I'm not entirely clear, but it sounds like you might have a field (#17) that
holds more than one fact (e.g., age AND height), and you're trying to create
a new field (#18) into which you'll put the second fact (e.g., height).

I don't understand about "delete the extra record".

It would help if you provide an example of the data. There's a chance your
data structure could benefit from further normalization. What's in your
fields?

Regards

Jeff Boyce
<Access MVP>
 
Back
Top