Table |
Description |
InputTable |
Contains items to cluster. |
SeedTable |
[Optional. Disallowed with OutputSeedsTable syntax element.] Contains seeds to use for hashing. Typically created by earlier MinHash call that specified its name with OutputSeedsTable syntax element. |
InputTable Schema
The table can have additional columns, but the function ignores them.
Column |
Data Type |
Description |
user_id_column
|
Any |
Identifier to cluster (for example, user identifier). |
item_id_column
|
BIGINT, INTEGER, or VARCHAR |
Identifiers of items on which to base clustering (for example, items that user purchased). Items are separated by delimiter. |
SeedTable Schema
Column |
Data Type |
Description |
index |
INTEGER |
Hash identifier. Values are from 0 to number_of_hash_functions - 1. |
a |
INTEGER |
Seed value that hash function used to create hash values that MinHash algorithm used. |
b |
INTEGER |
Seed value that hash function used to create hash values that MinHash algorithm used. |
p |
INTEGER |
Seed value that hash function used to create hash values that MinHash algorithm used. |