There's probably an easier way to do this than
the way I'm suggesting, but here's a go at it:
I assume that the dataset is in a range called
DATA, in column B, starting in B1. Prepare
three empty columns (I'll assume G, H, and I).
From G1 through Gx (where x is the number of
cells in the dataset), use the control-shift-enter formula
=IF(COUNTIF($B$1:B1,B1)=1,B1,"")
For sanity's sake, let's call G1 through Gx DATA2.
You can keep column G hidden.
H1 through Hx should have the (normal) formula
=IF(ISERR(SMALL(Data2,ROW(G1))),0,SMALL(Data2,ROW(G1)))
I1 through Ix should have the control-shift-enter
formula
=SUM(IF(H1=Data,1,0))/COUNT(Data)
Column G extracts unique entries from the
dataset. Column H sorts the data from smallest
to largest, and Column I does the math.
I've assumed, of course, that you have numeric
data. If you have alphabetic data, you should
check out Chip Pearson's page
http://www.cpearson.com/excel/noblanks.htm
to get the formula to use in Column H.
HTH
Thanks for your response. I'm now wondering if I should have provided a little more info. Thete are appr.
3000 entries in the one column. They are numeric, but they are a decimal figure.