Identifier-item file

See also: transaction data and attribute-value data, item list file, names file, data file.

An identifier-item file is a text file that lists transactions to be analyzed in identifier-item format.  Each transaction has a unique identifier.  Each line associates a transaction identifier with an item in that transaction.

Each line must contain at least two columns, one for the identifier and one for the item.

For identifier-item files, the Data Import Wizard:

• Allows the user to specify whether the data is ordered by identifier.  Ordered identifier-item files can be imported more efficiently than unordered files.
• Allows the user to select which columns contain the identifier and item.  Other columns are ignored.
• Allows the user to select the column delimiter.  Any character may be used as a column delimiter.
• Allows the user to sample from the data.  The user can select the percentage of transactions to import.  In this case a random selection of transactions is imported.  Only these transactions are used in the analysis.

For example, the following shows the content of an identifier-item file containing two columns delimited by commas, the first column containing the identifier and the second the item.

001, apples
001, oranges
001, bananas
002, apples
002, carrots
002, lettuce
002, tomatoes

Transaction 001 contains apples, oranges and bananas.  Transaction 002 contains apples, carrots, lettuce and tomatoes.