db normalization using table analyzer

[Guest]

Hi,

I've inherited a non-normalized db which I need to normalize. I thought I
would use the table analyzer to let Access determine how to split up
non-normalized tables into normalized ones. However, when I go through the
steps in the wizard, I come to a screen which says the following:

"The wizard has found some records with very similar values. Records
referring to the same information should be identical, so that they will be
combined into one unique record."

That makes sense as far as it goes, however, the records Access identifies
as having "very similar values" have anything but very similar values.

Any idea what's going on here? Is the table Analyzer wizard not useful?

 

[Guest]

Hi Dave,

I would caution against using it. The best database design is done with
pencil, a clean sheet of paper, and a good erasure. Don't trust the Table
Analyzer wizard to this important task!

Also, the table analyzer wizard will happily create table lookups for you in
some cases. Lookups defined at the table level art the creation of thy evil
one (see the 2nd commandment here:
http://www.mvps.org/access/tencommandments.htm


Tom Wickerath
Microsoft Access MVP

http://www.access.qbuilt.com/html/expert_contributors.html
http://www.access.qbuilt.com/html/search.html
__________________________________________

 

[Nick Coe \(UK\)]

It's not that the analyzer is untrustworthy, rather that it
cannot contain sufficient built in rules to analyse all sets
of circumstances. Use it as a first step.

In other words you have to apply some human experience and
intelligence to the output to get best results.

You sound experienced so you probably know that there are
times when it is not possible or even desirable to fully
normalise a database to 3NF. Some bits of it may need to be
de-normalised in order to work sensibly.

The upshot of my waffle is you take the results of the
wizard, don't treat them as gospel, and make such
corrections as you need to get a sane, working data model.

--
Nick Coe (UK)
http://www.alphacos.co.uk/




In Dave F typed:

 

[Nick Coe \(UK\)]

Tom,

My email reply to you is bouncing with a permanent error.

--
Nick Coe (UK)
http://www.alphacos.co.uk/




In Tom Wickerath typed:

 

[Guest]

Thanks for the responses.

You've basically confimred by suspicions.