Results 1 to 10 of 10

20060928, 19:16 #1
 Join Date
 Sep 2006
 Posts
 4
 Thanks
 0
 Thanked 0 Times in 0 Posts
determining the value of intersection (Excel/2003)
I have a question that has so far stumped myself and coworkers. I have 3 different trendlines plotted in an Excel chart. The lines intersect to form something looking like an H  so there are two different points of intersection.
How can I determine the values of those two intersecting points? They don't lie on any of the original points I plotted  they lie on the trendline. I would like to avoid doing this by hand, if possible.
Thanks in advance!

20060928, 19:34 #2
 Join Date
 Jan 2001
 Location
 Redcliff, Alberta, Canada
 Posts
 4,066
 Thanks
 2
 Thanked 5 Times in 5 Posts
Re: determining the value of intersection (Excel/2
Could you attach the file? I'm having trouble visualizing the problem.
[b]Catharine Richardson (WebGenii)
WebGenii Home Page
Moderator: Spreadsheets, Other MS Apps, Presentation Apps, Visual Basic for Apps, Windows Mobile

20060928, 19:36 #3
 Join Date
 Sep 2006
 Posts
 4
 Thanks
 0
 Thanked 0 Times in 0 Posts
Re: determining the value of intersection (Excel/2
Here it is!
 Katie

20060928, 19:47 #4
 Join Date
 Mar 2002
 Posts
 84,353
 Thanks
 0
 Thanked 29 Times in 29 Posts
Re: determining the value of intersection (Excel/2003)
Welcome to Woody's Lounge!
You can use the LINEST function to find the slope and intercept for each of the trend lines from the source data, then use straightforward maths to calculate the intersection points. See screenshot.
(You can also display the equations for the trend lines in the chart, by right clicking a trend line and selecting Format Trend Line, but then you'd still have to do the rest manually)
Post back if you need more help, preferably with a sample workbook.

20060928, 20:07 #5
 Join Date
 Sep 2006
 Posts
 4
 Thanks
 0
 Thanked 0 Times in 0 Posts
Re: determining the value of intersection (Excel/2003)
My lines are curves (I set the trendline to be a 2nd order polynomial)  is this a problem with using the LINEST or SLOPE function?

20060928, 20:10 #6
 Join Date
 Mar 2002
 Posts
 84,353
 Thanks
 0
 Thanked 29 Times in 29 Posts
Re: determining the value of intersection (Excel/2
Ah, you had already posted a workbook. Your trend lines are polynomials. It's still possible, though more complicated, to use LINEST to provide the equations for the trend lines. Calculating the intersections is more complicated too  see the attached version.

20060928, 20:19 #7
 Join Date
 Sep 2006
 Posts
 4
 Thanks
 0
 Thanked 0 Times in 0 Posts
Re: determining the value of intersection (Excel/2
Thanks! I really, really, really appreciate your help!
Katie

20060928, 22:37 #8
 Join Date
 Jul 2002
 Location
 Pittsburgh, Pennsylvania, USA
 Posts
 11,225
 Thanks
 14
 Thanked 342 Times in 335 Posts
Re: determining the value of intersection (Excel/2
To be a little more explicit in something Hans only hinted at:
Each of your curves could have more than one answer and you will have to choose which one you want or have some criteria to choose...
Steve

20060928, 22:52 #9
 Join Date
 Mar 2002
 Posts
 84,353
 Thanks
 0
 Thanked 29 Times in 29 Posts
Re: determining the value of intersection (Excel/2
And another point (for information only):
In this example, the trend lines are polynomials of degree 2. Calculating their intersection means solving a quadratic equation; this can be done using the famous (or notorious) quadratic formula shown below.
For polynomials of degree 3 and 4, you'd need to solve a cubic or quartic equation. There are explicit formulas for doing so, but they are horribly complicated. For higher degree equations, there simply is no explicit formula. In these cases, the solutions can only be approximated; Excel's Goal Seeker tool could be used for this.

20060929, 10:27 #10
 Join Date
 Jul 2002
 Location
 Pittsburgh, Pennsylvania, USA
 Posts
 11,225
 Thanks
 14
 Thanked 342 Times in 335 Posts
Re: determining the value of intersection (Excel/2003)
Is there a real theoretical relationship to the curves that you just fit to the quadratic. You might not get the extra and you may be able to solve them easier if you fit to a real relationship. Some relationships are parabolas, but not all curves are parabolic.
These equations, though most likely curved, can be fit relatively well with 3 simple lines and they would be much easier to solve...
Steve