Databases Reference
In-Depth Information
'(/(7()5200$5.(7B%$6.(7B7$%/(
,16(57,1720$5.(7B%$6.(7B7$%/(
,7(06(7B.(<
2%-(&7B.(<
48$17,7<
'2//$56
6(/(&7
,7(06(7B.(<
%+,(5$5&+<B.(<
68048$17,7<$648$17,7<B680
680'2//$56$6'2//$56B680
)520)$&7B7$%/($
,11(5-2,1
0$5.(7B%$6.(7B+,(5$5&+<B',0(16,21%
21$2%-(&7B.(< %2%-(&7B.(<
*5283%<
,7(06(7B.(<
%+,(5$5&+<B.(<
fIGure 7.8
Load Market Basket Table with hierarchies.
data in them. But, in general, the SQL in Figure 7.7 operates in a fashion
similar to the SQL in Figure 7.3 and Figure 7.5.
The output of the SQL in Figure 7.7 is shown in Table 7.10. The data in
Table 7.10 can be presented in the form of a view or a physical table for the
same criteria as discussed previously.
step 6: load a market Basket BI table using the
same hierarchy as Driver and correlation
If a Ma rket Basket Analysis effort can consider the correlations from
objects leading to hierarchical groups, and hierarchical groups leading to
objects, then that same Market Basket Analysis effort can also consider
the correlations from hierarchical groups leading to hierarchical groups.
It was inevitable. Using again the hierarchical groups in Table 7.6, the
SQL in Figure 7.8 extracts data from the original Fact table and loads the
MARKET_BASKET_TABLE.
This method works because the Driver Object and Correlation Object
are hierarchical peers. Contrary to the method of organizing data into
separate tables, this method shows that Market Basket Analysis methods
apply to all hierarchies of objects when the Driver and Correlation Objects
are hierarchical peers. When they are peers, no additional SQL is required
