Here’s More Help For A Huge Hardship
June 25, 2014 Ted Holt
|
In the April 16, 2014, issue of this august publication, I shared how encoded vector indexes can help with the task of summarizing data from large tables. Tom McKinley of IBM graciously emailed me with more information about that technique, and it is my pleasure to pass Tom’s wisdom along to you. Assume two tables–one of items, one of sales.
create table Items
( Item char (12) not null,
Description char (25),
Cost dec (7,2),
Price dec (7,2),
primary key (Item))
create table Sales
( ID integer as identity,
Invoice dec (9,0),
Line dec (3,0),
InvoiceDate date,
Customer dec (7,0),
|
