I need some help with develping some function... here what I need to
achieve:

Create "weights" for the 21 numbers listed below (in between ***s).

Their cumulative weight (sum) must equal 100% (or 1).

Here's the trick though... the weight of the 21 listed numbers below are NOT
sequencial.

For instance:

1.1 may equal = "0.2800"
1.2 may equal = "0.2200"
1.3 may equal = "0.1600"

but now,
2.1 may equal = "0.1000"
3.1 may equal = "0.0800"
4.1 may equal = "0.0500"

then

1.4 may equal = "0.0250"
1.5 may...
2.2 may...
3.2 may...
4.2 may...
1.6 may...
etc.


Essentially, there should be some relationship between the proportiate
counts within each of the 4 groups.

Group 1 contains: 12 records
Group 2 contains: 4 records
Group 3 contains: 2 records
Group 4 contains: 3 records

Again, the top numbers (e.g. "1.1 to 1.3") should get the greatest weights.
But the top number of the other 3 groups (2.1, 3.1, 4.1) must be greater
than the lower numbers of Group 1 (1.4 to 1.12).

At the same time, the lower numbers of Group 2 - 4 (e.g. 2.2 to 2.4; 3.2;
4.2 to 4.3) must be lower than the lower numbers of Group 1 (1.4 to 1.12).

Group 1 is the only group with a relative high count of numbers (compared to
the other 3 groups). So, maybe Group 1 should have something like "High
Numbers", "Medium Numbers", and "Low Numbers"... while the other 3 groups
have only "High and Low Numbers". I hope this make sense?

Lastly, this needs to be somehow dynamic... I may add or delete to/from
these 21 numbers. So, the function that calculates the "proportiate counts"
amongst the 4 groups must be automatically adjusted based on additions or
deletions.

This does not have to be 100% scientific-proof... nevertheless, I'd like
this to be as "scientific" as it can be.


****
1.1
1.2
1.3
1.4
1.5
1.6
1.7
1.8
1.9
1.1
1.11
1.12
2.1
2.2
2.3
2.4
3.1
3.2
4.1
4.2
4.3
****

Thanks for any ideas,
Tom