S
stratasmith
Here's what I need to do:
I have a rectangular array with values for a number of variable
(columns) in a number of samples (rows).
In another (square) array, I want to generate the variance of th
natural log of the ratios of all possible pairs of variables. Thus
cell mn of the output array has the variance of the logs of the ratio
of every pair of values in row m and row n of the input array. I ca
do this for an individual cell of the output array with an arra
formula, but as there are about 35 columns of variables there would b
over a thousand array formulas to type in!. I can't use Copy an
Paste, because of the way column n of the input becomes row n of th
output.
I could probably do it in a user defined function, but I'm not sure i
I can use an array formula in a udf, and someone may be able to thin
up a more efficient way of doing what I want.
Any ideas?
SS
I have a rectangular array with values for a number of variable
(columns) in a number of samples (rows).
In another (square) array, I want to generate the variance of th
natural log of the ratios of all possible pairs of variables. Thus
cell mn of the output array has the variance of the logs of the ratio
of every pair of values in row m and row n of the input array. I ca
do this for an individual cell of the output array with an arra
formula, but as there are about 35 columns of variables there would b
over a thousand array formulas to type in!. I can't use Copy an
Paste, because of the way column n of the input becomes row n of th
output.
I could probably do it in a user defined function, but I'm not sure i
I can use an array formula in a udf, and someone may be able to thin
up a more efficient way of doing what I want.
Any ideas?
SS