Can you help me find the equation from this graph?

[strokebow]

Finding an Equation from a Graph

Hi,

I need to find an equation from a graph.

The graph itself is not very complex. Just need some help.

I can't really show you the graph but I'll give you the points so you can plot it to paste them into Excel or something.

x co-ords y co-ords
0 0
1 1.5
2 3
3 3
4 3
5 1.5
6 0
7 0
8 0
9 -1
10 -2
11 -2
12 -2
13 -1.5
14 -1
15 -0.5
16 0

Please help . . . your help is much appreciated

Thanks

 

[strokebow]

Hi,

Well this isn't really homework - Its just buggin me how I would do this.

I want to find an equation from a graph.

The graph itself is not very complex. Just need some help.

I can't really show you the graph but I'll give you the points so you can plot it to paste them into Excel or something.

x co-ords y co-ords
0 0
1 1.5
2 3
3 3
4 3
5 1.5
6 0
7 0
8 0
9 -1
10 -2
11 -2
12 -2
13 -1.5
14 -1
15 -0.5
16 0

Please help . . . your help is much appreciated

Thanks

 

[strokebow]

Finding Equation from Graph

Hi,

Yeahhh I just posted this in the wrong forum.
Anyway this isn't homework - Its just buggin me how I would do this.

I want to find an equation from a graph.

The graph itself is not very complex. Just need some help.

I can't really show you the graph but I'll give you the points so you can plot it to paste them into Excel or something.

x co-ords
0
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16

corresponding y co-ords
0
1.5
3
3
3
1.5
0
0
0
-1
-2
-2
-2
-1.5
-1
-0.5
0
Please help . . . your help is much appreciated

Thanks

 

[Maxwells Demon]

what's the problem? You just plot them into a CAS-tool and use regression of some kind? and then you'll get x as a function of y..

 

[StatusX]

There are many, many equations whose graphs would pass through those points, and even more which would come close to them. Exactly what type of equation are you looking for (polynomial, exponential, etc)?

 

[strokebow]

Hey thanks for your reply. Whats a CAS tool? Like what?
Im not sure I have the right software for it. Is there any manual way to do this?


Thanks

 

[strokebow]

I just want to relate x and y over those values via function
How would i Go about doing this?

 

[Maxwells Demon]

it's like an advanced calculator like a Texas Instruments Ti-89 Titanium like mine, or you could use Excel like you say..



I'm from Denmark and use the Danish excel so I'm not sure what the different menus and functions is called in English..

not exactly a "pocket" calculator, but really good..

 

[StatusX]

Think of it like this. Say you find a function f that passes through all of those points. Now think of another function g that is zero at x=0,1,...,16. You should be able to find a lot of these (eg, g=0, or g=sin(pi*x), or g=x(x-1)(x-2)...(x-16), or g=x(x-1)(x-2)...(x-16)(x-123), etc). Then adding f to g gives another function that passes through all the points. So, like I said, there are a lot of them. What is this for?

 

[arildno]

There are an infinite number of continuous functions whose graphs go through all of those points.

that is, your question is ill-posed.

 

[arildno]

Do not double post your questions.

 

[ssd]

What do you want: a curve passing through all the points or a curve to substitute the function from which the points were derived with some possible errors in obsrevation.

 

[ZapperZ]

2 threads have been merged because the OP made multiple posts. So if the flow makes no sense, don't blame me.

Zz.

 

[strokebow]

basically its some values I determined experimentally in the labs. And over 16secs the graph is drawn. Now, I'm using a program to simulate my findings but I need an equation to relate this graph. Can anyone give me an equation ofr tell me how I can work one out?

 

[strokebow]

Hey thanks for replies

basically its some values I determined experimentally in the labs. And over 16secs the graph is drawn. Now, I'm using a program to simulate my findings but I need an equation to relate this graph. Can anyone give me an equation ofr tell me how I can work one out?

 

[Maxwells Demon]

I would suggest that you plot them into a coordinate-system using Excel.. Then looking at them and choosing the right kind of regression (power, quad, lin, exp and so on)...

Make an experiment and see which regression brings the R^2 closest to 1...

 

[strokebow]

Thanks for your replies!

But you aint really helped. Is there anyone who can give me a simple explanation how to do this or who can actually do it and walk the walk.

No offence I'm thankful for your efforts but I need telling simply because I am rather simple myself.


thanks

 

[StatusX]

Do you understand what we mean when we say there are many equations whose graphs touch those points? There's no single "equation for this graph". So unless you're more specific about what you want, we can't help you.

 

[Maxwells Demon]

As i said make an experiment and see which regression brings the R^2 closest to 1...

 

[Integral]

The only way we could help you arrive at a meaningful function discription is if you tell us what the physcal system these measurments come from. That way we couldmodle the system and not the data.

 

[D H]

You can get a perfect fit (R=1) with a high-enough order polynomial. One problem: As your measurements are not perfect, that perfect-fit polynomial is almost certainly wrong. What you want to do instead is find a fit to some reasonable model, and then test to determine if the fit is reasonably good.

 

[Maxwells Demon]

D H that's also an (better) opportunity...

 

[strokebow]

Basically,

Well it's for the speed of an object. Although, this specific data isnt. The data I have provided for you is from a textbook. ITs from an old textbook where the reader is asked to derive the equation from the graph (the one I gave you). You are to derive an equation for that graph over 16seconds. Thats what I wanted to know how to do and if it were not possible for anyone to tell me how to do it I would like to know what the equation is because it has been bugging me.


Thanks again. I value your replies and taking time to read my thread.

thanks

 

[mherna48]

I have the same problem. I have over a thousand data points and none of the trendlines from excel fit the line. Any other ideas?

 

[lilchicadees]

I need help. I need to find an equation to the attached graph.
So far i have y=3sin(pi/2x).
I guess that is wrong, since the period is wrong.
So I need to figure out 4pi = 2pi/b, b=?
Please help!