trendlines and their equations

  • Thread starter Thread starter Jeff
  • Start date Start date
J

Jeff

I've recently encountered a problem where the trendline
equation doesn't match its corresponding trendline
depicted on the graph. Has anybody else seen this?
 
Jeff said:
I've recently encountered a problem where the trendline
equation doesn't match its corresponding trendline
depicted on the graph. Has anybody else seen this?
Click to expand...

How could you tell? What was the fitted trendline's equation? Did the graph
have any extreme or inflection points? How many data points were fitted?
 
Need more information to ID what the issue is.

The trendline equation displayed on a graph is heavily rounded by
default. Right click on it and format to Scientific notation with 14
decimal places, to copy values for computation.

The equation computed by LINEST may not agree with the one displayed on
the graph if the problem is ill-conditioned (numerically difficult); in
that case the computation of the graph equation is numerically better.

Jerry
 
Back
Top