>>>Hi Dragan,
>>>
>>>Thanks for the heads up. I've been wondering how much of an increase in performance in queries could be gained by converting numeric values (where appropriate) to integer and currency types. Haven't done any testing yet, but I will. It would seem to me that using these data types could yield significant improvement, but that's a SWAG also.:-)
>>
>>How would one convert numeric to integer or currency within a query? Any difference between doing this with native data vice ODBC data?
>
>Hi Cindy,
>
>I'm afraid you misunderstood. I was speaking of converting the underlying table structures. My feeling is that by changing the type from numeric, you would eliminate the conversion of the underlying data (which is stored as character actually) and thus speed up any queries using aggregate functions. This could be significant with queries that had to SUM(), for example, a large number of rows into a smaller subset.
Thanks for explaining George. I will be using integer and currency every chance I get.
That is so totally weird that numerics are stored as characters, though I guess that's the same as what's in an ASCII file! Here's a weirdie for you: I retrieved data the other day from an Oracle RDB database (not "regular" Oracle) and the field was supposed to be numeric, but came across as character! I had to SELECT MyNumericField + 0.0 to get it to come across as numeric!
Précédent
Suivant
Répondre
Voir le fil de ce thread
Voir le fil de ce thread à partir de ce message seulement
Voir tous les messages de ce thread
Voir tous les messages de ce thread à partir de ce message seulement