Logarithmic trendline y-intercept?

[RiverRocks]

Hi,

I have an X,Y scatter plot and I've fit a logarithmic trendline to the data. The nature of the data is that every X=1, Y=100%; however, the trendline does not pass through this as the y-intercept, nor does "b=1" in my equation of y=-(a)ln(x) +b. Is there a way to set the y-intercept as X=1, Y=100%? In the Trendline options, the "Set Intercept" boxed is grayed out. I realize this may not give the best fit trend line anymore, but due to the nature of the data I think the y-intercept should be set at X=1, Y=100. Sorry if this is confusing, I'm not in mathematics or statistics.

LogTrend1.JPG

Thanks for any help!

[shg]

You could use Solver to get a best-fit value of m for y = m ln(x) + 1

[RiverRocks]

I've never used Solver before, would you be able to explain how to do that? Thanks!

[shg]

Post a workbook and I'll add a Solver model.

[RiverRocks]

OK here is a workbook with the chart and data from photo above. Thanks very much for your help!

LogEqn_Trendline_v1.0.xlsx

[shg]

See attached.

Frontline Systems has lots of material. You could start at http://www.solver.com/optimization-tutorial

[shg]

In response to your PM: You're welcome.

In terms of getting responses to future questions in the forum, it's better to say thank you here, so that people see that you acknowledge assistance.

Kind regards,