Updating reports (1 Viewer)

[bigcraigie]

I have to import from excel updated info 2 or 3 times a day.

Client Latest Report

A Bloggs 01/02/12
B Smith 01/02/12

So if I update the report from excel, and run the report based on the query, how can I get it to run the report and show only the latest report?

ie

A Bloggs 02/02/12
B Smith 01/02/12

where the table will still show both Bloggs entries - sorry quite new to this and I know it might be quite simple but hope to have it running by Friday.
Thanks in advance

 

[vbaInet]

Welcome to the forum!

Do you have multiple records per client, such as below?

Client Latest Report

A Bloggs 01/02/12
A Bloggs 31/01/12
A Bloggs 02/02/12
B Smith 01/02/12

Do you have a Client ID field?

 

[bigcraigie]

Hi, thanks!
Yes the table will build up records until the client stops reporting.
I have the 'Name' column as the primary key - but do I have to introduce a 'Client ID field'?
if it is any help the client names will be unique throughout their history.
BigC

 

[vbaInet]

How can you guarantee that the names will be unique? Besides that point normalization requires you give it a unique numerical ID. Do you know about normalization?

In the meantime, if you create a Totals query GROUPED BY Client Name and Max() on the Date field, it should give you what you want.

 

[dportas]

How can you guarantee that the names will be unique? Besides that point normalization requires you give it a unique numerical ID. Do you know about normalization?

Normalization does not require a "unique numerical ID". Normalization requires that every table have at least one key and normalization theory specifies certain rules about dependencies on those keys. It says nothing about what type of data the keys should contain.

Of course it's a good thing to think carefully about how to identify things in the database, the type and design of keys and what constraints to implement - but the idea that you need to invent numbers for things has nothing at all to do with normalization.

 

[vbaInet]

Normalization does not require a "unique numerical ID". Normalization requires that every table have at least one key and normalization theory specifies certain rules about dependencies on those keys. It says nothing about what type of data the keys should contain.

I was speaking with respect to the OP's specific structure. A numerical ID is much more effecient than an ID of type String.

 

[dportas]

I was speaking with respect to the OP's specific structure. A numerical ID is much more effecient than an ID of type String.

It may be more efficient or maybe not. But what you said was that normalization required it and I thought that point was worth correcting.

 

[vbaInet]

It may be more efficient or maybe not. But what you said was that normalization required it and I thought that point was worth correcting.

Like I said earlier, that was in the context of the OP's specific problem. I wasn't explaining what normalization is, I was pointing that a numerical ID would be more appropriate in his/her case to satisfy normalization.

It is also common knowledge that a numerical identifier is much more effecient than a text identifier.