6.2. Sample Crosstab Report

To create a crosstab report, select Crosstab band orientation on the Report structure tab of the report editor. This orientation automatically adds three datasets to the band:

  1. <band_name>_dynamic_header – the data from this dataset is copied to the right like a vertical band with table columns header.

  2. <band_name>_master_data – the data from this dataset is copied downwards like a horizontal band with table rows header.

  3. <band_name> – the dataset named the same as the band, it belongs to. It is the main content band that will fulfill the matrix cells.

These datasets can have any of available dataset types: SQL, JPQL, Groovy, and so on.

For example, a crosstab report for the Order entity of the Sales sample application has the following structure:

crosstab structure
Figure 62. Crosstab Report
  • Here, the orders_dynamic_header dataset will return the list of months names:

    orders_dynamic_header dataset
    import java.text.DateFormatSymbols
    
    List result = new ArrayList()
    DateFormatSymbols dateFormatSymbols = DateFormatSymbols.getInstance(Locale.ENGLISH)
    for (i in 0..dateFormatSymbols.months.length - 1) {
        result.add(["header_id" : i + 1, "month_name" : dateFormatSymbols.months[i]])
    }
    return result
  • The orders_master_data dataset returns names and identifiers of the customers selected by the user as an external report parameter:

    orders_master_data dataset
    select name as name, id as customer_id
    from SALES_CUSTOMER
    where id in (${selected_customers})
  • The orders dataset will provide data for the matrix cells, which is the order amounts sum for the particular month and customer. It takes orders_master_data@customer_id (the customer id) as the Y-coordinate of the cell, and orders_dynamic_header@header_id (the month name) as the X-coordinate, and fills the matrix cell with the amount value.

    In the example below the report has two more external parameters: the start_date and the end_date that define the range of order dates. It would be a good idea to set up cross-parameter validation to make sure the range makes sense.

    orders dataset
    select
            o.customer_id as orders_master_data@customer_id,
            month(o.date_) as orders_dynamic_header@header_id,
            sum(o.amount) as "amount"
    from sales_order o
    where o.date_ >= ${start_date} and o.date_ <= ${end_date}
    and o.customer_id in (${orders_master_data@customer_id})
    and month(o.date_) in (${orders_dynamic_header@header_id})
    group by o.customer_id, month(o.date_)
    order by o.customer_id, month(o.date_)

Then you can create a report template using Microsoft Office or LibreOffice.

The report template should contain named regions for all three datasets of the crosstab band and, additionally, the named region for the column header: <band_name>_header. In our case it is orders_header.

Below is an example of a template which outputs the list of Customers vertically and Orders made by each customer horizontally grouped by the months the orders were placed.

crosstab template 2
Figure 63. Crosstab template
crosstab names regions
Figure 64. Named regions

As a result, the report is extended both vertically and horizontally and aggregates the order amounts for each customer and each month:

crosstab result
Figure 65. Crosstab result

If you want to add totals for the report, you should do it in separate bands with their own datasets.