If by RMS you mean RMSE, the estimated standard deviation about the
regression line, the sum of squares for error would be
=SUMSQ(y-x)
with n-1 degrees of freedom for an assumed model of of y=x. This is an
array formula, that must be array entered (Ctrl-Shift-Enter). Thus the
RMSE would be
=SQRT(SUMSQ(y-x)/(COUNT(y)-1))
which also must be array entered.

Jerry

aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa wrote:

by "Line of equality" I mean the line x=y. I have the regression set (x, y)
and I want the RMS considering that the fit line is the line of equality. Is
it possible?

"Jerry W. Lewis" wrote:


What do you mean by "line of equality"?

Jerry

aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa wrote:


Jerry,

thank you for your response. You have right, it is a regression question but
I need the RMS for the line of equality and not for the trend line. Is the
function you proposed the proper one for my case?

"Jerry W. Lewis" wrote:



Thank you for providing an informative subject line, but it usually
helps to elaborate in the body of the post. As Earl noted, the most
obvious elaboration would be RMS of what?

If this is a regression question, you would use
=INDEX(LINEST(yData,xData,const,TRUE),3,2)

Jerry