View Single Post
  #6   Report Post  
Harlan Grove
 
Posts: n/a
Default

Brad wrote...
I have the following data in the left-most column and the results of a
formula in the next:

....
The formula normalizes the first column data so that the values in the
second column are valued between 1 and 0. The formula is:

=(A131-SMALL($A$131:$A$145,1))/(LARGE($A$131:$A$145,1)-SMALL($A$131:$A$145,1))

My problem is that, in the event that there is a zero value in the column of
numbers (the numbers almost always include positive and negative values), I
would like the normalizing formula to show a zero value in the results
there. Note that the actual 0 datum when normalized through this range,
equals .571774.


What are you trying to accomplish by this? If your data values span
positive and negative values, then they must also span zero values
unless the underlying process is discontinuous at zero, but in that
case you should never get a zero data value.

Since the zero value would normally not be the middle value of the range of
numbers, I think if I could come up with a formula to normalize all the
values greater than 0 with a range of 0 to 1, then all negative numbers 0
to -1, and 0= 0, I'd have what I am looking for. . . .


I'd guess this means that you're unconcerned about the relative
absolute values of the averages of positive and negative values, e.g.,
if positive values spanned 0.05 to 1.8 while negative values spanned
-0.001 to -0.080, you'd be unconcerned that normalized 1.0 corresponded
to original 1.8 while normalized -1.0 corresponded to -0.080. If so,
try the array formula

=A131/MAX(IF(SIGN($A$131:$A$145)=SIGN(A131),ABS($A$131:$ A$145)),--(A131=0))