[AccessD] Fastest Way

Rocky,

I would suggest the following approach.

1/ Import all of the spreadsheet data to a temporary table.
2/ Create a query joining the import table to the table tblPatientTest using all five fields. This query will list records already present in tblPatientTest. For example (using my field names)...

SELECT tblImports.Policy_ID, tblImports.Seller_ID, tblImports.dSubmissionReceived, tblImports.dAcknowlegementSent, tblImports.PolicyType
FROM tblPatientTest INNER JOIN tblImports ON (tblPatientTest.PolicyType = tblImports.PolicyType) AND (tblPatientTest.dAcknowlegementSent = tblImports.dAcknowlegementSent) AND (tblPatientTest.dSubmissionReceived = tblImports.dSubmissionReceived) AND (tblPatientTest.Seller_ID = tblImports.Seller_ID) AND (tblPatientTest.Policy_ID = tblImports.Policy_ID);


3/ Create another query, an Unmatched Query, which will be the same as the first query, except you use RIGHT joins instead of equi-joins. Like so...

SELECT tblImports.Policy_ID, tblImports.Seller_ID, tblImports.dSubmissionReceived, tblImports.dAcknowlegementSent, tblImports.PolicyType
FROM tblPatientTest RIGHT JOIN tblImports ON (tblPatientTest.PolicyType = tblImports.PolicyType) AND (tblPatientTest.dAcknowlegementSent = tblImports.dAcknowlegementSent) AND (tblPatientTest.dSubmissionReceived = tblImports.dSubmissionReceived) AND (tblPatientTest.Seller_ID = tblImports.Seller_ID) AND (tblPatientTest.Policy_ID = tblImports.Policy_ID)
WHERE (((tblPatientTest.Policy_ID) Is Null));

That will list just the data that is different in one or other of the five fields. (The principle of unmatched queries is simple. You right join on all the fields you want to check... RIGHT JOIN tblImports ON (tblPatientTest.PolicyType = tblImports.PolicyType)... And then select a field from the table that is not supposed to contain the data. It that field in null then the new data is not present in the destination table ...WHERE (((tblPatientTest.Policy_ID) Is Null));

I ran this query against a table with 12,000 records and it more or less instantly returned the 189 new rows that I had engineered in some test data.

HTH

Lambert

-----Original Message-----
From: accessd-bounces at databaseadvisors.com [mailto:accessd-bounces at databaseadvisors.com] On Behalf Of Rocky Smolin
Sent: Thursday, January 14, 2010 10:01 AM
To: 'Access Developers discussion and problem solving'
Subject: [AccessD] Fastest Way

Dear List:

I have a medical database into which lab tests are imported from a spreadsheet (exported from a mainframe medical records database at a big hospital).  Front and back ends are Access 2003. The table tblPatientTest has (in addition to the autonumber PK) 5 fields:

Patient ID (FK)

Test ID (FK)

Test Date as Date/Time (indexed duplicates OK)

Test Time as long integer - value between 0000 and 2399 (indexed duplicates
OK)

Test Result - Text (indexed duplicates OK)

I'm using DAO to add the records to the table.  Imports will happen monthly and the number of records will be on the order of 10,000-100,000 with mean probably around 20k-40k.

As the tests are imported I want to check the table to see if the test record is already present by finding a record which matches all 5 fields to prevent duplicate test records from being entered.

Since the import will take place monthly, import time is not super critical but before importing the user will run an Error Report which does all the error checking that the import does without actually doing the import.  The user then resolves the errors and when the Error report is clean, will run the import.  So the error report may have to be run more than once to clean the data.

It is possible that the user will try to import a spreadsheet that has already been imported. So it's important to screen for duplicates. 

So I want to find the fastest method of checking for duplicate.

The obvious thing is rs.FindFirst on all five fields, but I'm thinking that may take a really long time.

Another possibility would be to create a temporary table in the front end or external mdb, and create indexed records by concatenating all five fields and checking for dups on that table.  Probably faster than the first method but creating the table could take a lot of time when the Patient Test table has a few hundred thousand records in it.

Or I could add the field of concatenated fields to the table itself as the records are added, indexed and maintain that.

Or some other method that I haven't thought of or don't know about.

Any advice welcome.

MTIA

 

Rocky Smolin

Beach Access Software

858-259-4334

www.e-z-mrp.com <http://www.e-z-mrp.com/> 

www.bchacc.com <http://www.bchacc.com/>