Exponential Regression using LOGEST

Question

I have two columns of data:

x	y
1	0.8
1200	0.5
3000	0.25
18000	0.11
45000	0.08
150000	0.05

I would like to use the LOGEST function to calculate y-values for a new set of x-values, such as:

x
1
500
1000
15000
50000

Unfortunately, just using the LOGEST function on the two columns ("=LOGEST(B2:B7,A2:A7,FALSE,FALSE)"), I get

0.99997555	1

This does not even fit the data points I already have.

Answer 1

(Resubmitted with a major correction based on the original requirements)

@NewerOfficeUser, please refer to my response to HansV.

As I noted, you can use a single-valued GROWTH function to estimate y for new values of x.

But as I demonstrate in the image in that response, the exponential regression of the form y = m^x is a poor estimate of your data.

Is there a reason why you are fixated on that regression formula?

IMHO, a power regression formula of the form y = x^m is a much better fit.

Converting y = x^m to the linear formula ln(y) = m*ln(x) , we can use LINEST(LN(B3:B8), LN(A3:A8), FALSE) to derive m, as demonstrated in E3:F3.

(FYI, more generally, F3 is actually ln(b). We can derive b with =EXP(F3) in G3.)

Then the formula for est y = x^m is =A3^$E$3 in C3, which we can copy into C4:C8 and C11:C15.

Download the Excel file (click here) for details. Ignore any preview errors and request to log in.

Aside.... In my version of Excel (YMMV), we cannot add a power trendline with intercept=1. So I created the equivalent by adding a chart series (in red) based on the values in C3:C8.