Excluding dimensions but NOT selections in a chart expression

[#1]

Hi,

This might be a very basic problem, but I've found no way of working around it yet.

For four different lengths of treatment (for a disease), '1 to 7 days', '8 - 14 days', '15 - 21 days', '22 - 28 days', I have recorded the numbers of males and females that were involved in each. I also have recorded for each participant an age segmentation, eg. '15 - 20' or '20 - 25'.

What I would like to show is a bar graph displaying, for males AND females (two separate series) the %ages, out of the selection, that occupy each age segmentation. The selection, in this case, is the length of treatment as described above.

So essentially I am looking to make a bar graph, showing %ages against age segmentation, with two series, male and female, with a list box containing each length of treatment. When each length of treatment is selected/deselected accordingly, the %ages alter.

The only way I can see of doing this (with my limited knowledge) is to involve, in each expression, a metric calculating the total number of participants that includes the selection but EXCLUDES the dimension, since I am calculating the %age out of the count over all the age brackets.

I know that you can exclude the current selection with {1<, and also exclude BOTH dimension and selection, but I can't find a way of excluding the dimension from the metric but including the current selection. Is this impossible, or is there a simple (or not simple) way of doing this?

Thanks to anyone who can help me on this.

[#2]

You can exclude dimensions in a aggregate function using the TOTAL modifier.

Sum(TOTAL Sales)

If you want to exclude most dimensions, but include one, you can use:

Sum(TOTAL <Respect> Sales)

Both of those functions will respect the selections.

[#3]

try using this:

SUM({<dimension_name={'dimension_value'}>} metric)