‎2007 Apr 10 8:53 AM
Hi,
i have a project where i need to work with 1, 5 million of datasets from a table. what would be the most effective way to do that assuming i have to loop at the table and compare every row to all the others. Would internal table still be suitable ?
Clemens
‎2007 Apr 10 9:00 AM
hi,
Internal table only will be a better choice ...
Regards,
Santosh
‎2007 Apr 10 8:55 AM
Hi Clemens,
Could you please elaborate on the requirement ..it would be helpful to suggest.
Regards
Mr Kapadia
‎2007 Apr 10 9:00 AM
Hi,
what i want to do is check wheter a table ( e.g. of debitors ) contains duplicates
( produced by spelling-mistakes etc. ) - as i have to go trough the whole table for that and compare every debitor to all the others i would like to know, wheter this can be done well with an internal table or if i should go for something more efficient.
regards,
clemens
‎2007 Apr 10 12:11 PM
As you will want to edit the name, (to remove spaces), before you sort the list you will need to use an internal table.
If you run out of space you could use an extract dataset, read about FIELD-GROUPS and EXTRACT in the help.
MattG.
‎2007 Apr 10 9:00 AM
hi,
Internal table only will be a better choice ...
Regards,
Santosh
‎2007 Apr 10 9:30 AM
Hi,
Internal Table is Better choice.
To avoid duplicate entries use the DISTINCT addition in the select query when filling up the internal table.
Use where conditions whenever poosible, read only the required rows or columns into the internal table.
Donot let the internal table become too large: use appropriate Table Type.
->Standard table for multiple access types; use SORT / BINARY SEARCH for mass accesses, and try to SORT only once
->Sorted table if generic key access is the main access type
->Hashed table if single line access with fully specified key is the only access type
Try and use APPEND as much as possible, instead of INSERT.
Make all changes in the internal table and block update into the database table.
Use the TRANSPORTING addition when modifying the internal table.
Hope U did all of them.
Award pts for useful answers
regards,
Naveenan.
‎2007 Apr 10 12:35 PM
hi
good
i dont think you would face any problem by using the internal table to store any amount iof data.
You can able to use the selct query to select as many number of files..
thanks
mrutyun^
‎2007 Apr 11 10:24 AM
Yes, you would have to use an internal table if you would want to compare every record against a given criteria. The only problem is the peformance of your program since you are saying you need to process 1.5 million records.
Hope this helps...
P.S. Please award points if it helps...