drop table Orders, Salesmen, T456a;
|
This line results in three tables being dropped from memory. |
Tab1:
Load * Inline [
Customer, Items, UnitPrice
Bob, 5, 1.50
];
Tab2:
LOAD Customer, Sum( Items * UnitPrice ) as Sales
resident Tab1
group by Customer;
drop table Tab1;
|
Once the table Tab2 is created, the table Tab1 is dropped.
|
// Load mapping table of country codes:
map1:
mapping LOAD *
Inline [
CCode, Country
Sw, Sweden
Dk, Denmark
No, Norway
] ;
// Load list of salesmen, mapping country code to country
// If the country code is not in the mapping table, put Rest of the world
Salespersons:
LOAD *,
ApplyMap('map1', CCode,'Rest of the world') As Country
Inline [
CCode, Salesperson
Sw, John
Sw, Mary
Sw, Per
Dk, Preben
Dk, Olle
No, Ole
Sf, Risttu] ;
// We don't need the CCode anymore
Drop Field 'CCode';
Drop Mapping Table map1;
Metrics:
Load * Inline [
Transaction ID, Sales Amount, Sales Cost, City, Customer ID
1000012, 4509.33, 1234.22, Oslo, 250v9849
1000013, 1043.21, 180.23, Gothenburg, 195d5930
1000014, 6038.35, 1400.10, Copenhagen, 195d5930
];
|
This example shows how the Drop mapping table variant can be used. A mapping table map1 is created and then used in a subsequent LOAD statement with the ApplyMap function. After the mapping table has been used, it is dropped from memory, optimizing RAM usage for the data to be loaded in subsequent tables of the script execution.
In this case, with only a few rows in the mapping table, the RAM optimization would be minimal. However, for large mapping tables or when creating mapping tables within loops, performance can be significantly improved.
|