Assigning exponential weights depending upon sample size

[64jnk]

I'm trying to find a function which will assign exponential weights depending upon sample size.

nu=an equally space coefficient sequence (.05,..,.5; by=.05).
m=sample size (10 in this case)

Adding each observation in nu, implies a weight of 1, which makes the sum of weights m.

I need to the exponential weights to add to m, given I'd like to compare the sum of this exponentially weighted vector to the sum of a uniformly weighted vector.

Would anyone be able to help with this please?

Many thanks,

 

[Stephen Tashi]

I need to the exponential weights to add to m

Pick m initial exponential weights a (whatever you mean by that), add them up. Let the sum be S. Then let the final weights be w = a (m / S).

If that doesn't suit you, try explaining what you mean by "exponential weights".

 

[64jnk]

I'm slightly embarrassed by how obvious this answer was. Thanks Stephen.