Group Help

[AvgMoJoe]

Crystal 2008
MS SQL DB

Given the following data set as it comes from my query:

I am using formulas for grouping since I am reporting differently for different Xtypes.

For Xtypes 2, I'm grouping by MeasureValues indicated by Measure_ID 1600, but I need to also include those 2's with no MeasureValues indicated by a Measure_ID 1600 as "*Unspecified*".
Furthermore, I have to be careful that 2's with 1600's don't also show up on my "*Unspecified*" group where the Measure_ID differs from 1600.

My end result (without the comments) should look similar to this:

2's[tab]<value>[tab]<count>
[tab][tab]goose[tab][tab][tab]2[tab][tab][tab]// (John and Fred)
[tab][tab]chipmunk[tab][tab]1[tab][tab][tab]// (Tom)
[tab][tab]*Unspecified*[tab]2[tab][tab][tab]// (Max and Arnold)

3's[tab]<value>[tab]<count>
[tab][tab]apple[tab][tab][tab]1[tab][tab][tab]// (Dave)
[tab][tab]pear[tab][tab][tab]1[tab][tab][tab]// (Paul)

I'm having trouble including the unspecified group without including people with Measure_ID 1600 in it (because they have other Measure_ID's). I'm not sure if this is a good case to use Group Selects or formulas or whatnot....

 

[Charliy]

if {table1.xtype} = 2
and (isnull({table3.measurevalue}) or {table3.measurevalue} = ' ')
then "*Unspecified"
else
If {table1.xtype} = 2
then {table3.measurevalue}
else
if {table1.xtype} = 3
then {table4.othervalue}

 

[AvgMoJoe]

Assume "John" and "Tom" are unique ID's. (Obviously I tried to oversimplify my real data )
(Besides not filtering for Table2.Measure_ID = 1600), this would cause John (1200 fox and 1700 wand) and Tom (1300 bear) to also be counted as "*Unspecified*", but they really have been specified as goose and chipmunk (respectively).