Currently I have a dataset which contains two tables: Loads and Maps. Each load will contain at least one map but could contain more. I'm using a relational db schem where the load has a row ID (RID) and each map entry has a RID that is used to associate it to a load. My XML looks like this:

Code:
  <Loads>
    <desc>This is my Description</desc>
    <options>-noio</options>
    <loaded>0</loaded>
    <rid>1</rid>
  </Loads>
  <Map>
    <rid>1</rid>
    <drive>S</drive>
    <path>Some path</path>
  </Map>
<Map>
    <rid>1</rid>
    <drive>R</drive>
    <path>Some other path</path>
  </Map>
This works but I would like to have the map data embedded in the Loads table:

Code:
<Loads>
    <desc>This is my Description</desc>
    <options>-noio</options>
    <loaded>0</loaded>
    <Maps>
        <Map>
            <drive>S</drive>
            <path>Some path</path>
        </Map>
        <Map>
            <drive>R</drive>
            <path>Some other path</path>
        </Map>
    </Maps>
</Loads>
I'm using ds.writexml to populate the xml files. How would I organize the dataset to generate the example above?