When generating a report that contains a cross tab, BIRT creates one column or one row in the cross tab for each unique value in the dimensions that you insert in the cross tab. If a cross tab displays sales totals by products in the column area and months of a particular year in the row area, and there are 100 products in the data set, the cross tab consists of at least 100 columns, 12 rows and 1200 cells, not counting a row and column for the grand totals. This cross tab is obviously too wide to view on‑screen or print. You can reverse the dimensions so that the product values appear as rows, and the month values as columns, but then the cross tab is much taller.

Although there is no limit to the amount of data that a cross tab can display, limiting the amount improves the usability of the data. After all, the main advantage of presenting data in a cross tab is the ability to compare and analyze information, preferably on a single page or screen.

As with any other type of report, you narrow the scope of data to include in a cross tab by creating filter conditions. If the data originates from a database, it is best, for performance reasons, to specify the filter condition in the SQL query. Alternatively, filter at the data set level, the cube level, and finally, at the cross tab level. The filtering option you select depends on whether other report elements or cross tabs are using the same data set or cube.