curve fitting for different data

Hello,

I have a data from different factors, I need to make a fitting curve as an estimation function for them.

[Star Strider]

Do you have a mathematical description of the process that created these data? Any number of functions could fit them, however estimating the parameters of the process that created them is the only one that would make any sense.

[Ahmed Abdalazeez]

Hi,

Thank you for your response!

I normalized these data by taking the parameter-constant- (L/\lambda)^1/4, so then I get thier distribution as it shown in this figure. Now I want indepentent function can estimate y according to x.

[Star Strider]

My pleasure.

It would be extremely helpful to know exactly what you are plotting, and the mathematical model of the process that produced those data, as well as the normalisation constants and other information.

We still do not have any of that information.

[Ahmed Abdalazeez]

Thanks a lot!

Well, Y- axces is the runup values times (Lnegth of beach slope/Wavelength = 20/70) vs steepness of incident wave. The data isa results of NLSW equations. Both xand y nondimensional. These different colors are different intial periods.

Hope these information help you to get helping from you :)))

Thanks in advance!!

[Star Strider]

What parameters would you want from a curve fit to your data?

Are there any literature references that you can post (preferably as .pdf files) that could explain this in more detail, for those of us who (like me) have very little understanding of littoral dynamics?

Have others estimated these parameters? If so, what models did they use?

[Ahmed Abdalazeez]

Actually, I am just looking for the best power curve fitting of (s/s0), I found:

fit= ((70/20).^(1/8)*(((s/s0).^(1/4)))); but I am not sure if it is the best fitting.

[Star Strider]

What parts of this equation:

fit  = ((70/20).^(1/8)*(((s/s0).^(1/4))));

represent parameters you want to fit? The exponents? s0?

[Ahmed Abdalazeez]

I need the fitting of (s/s0) in x axces

[Star Strider]

As I interpret it, ‘s’ are data, so ‘s0’ would be the parameter you want to estimate. Is this correct?

Before I attempt to extract your data from the figure you posted, what are they? If you already assumed a value for ‘s0’, then there is nothing to estimate.

I am still not certain what you are doing, or what your data are.

[Ahmed Abdalazeez]

No, (s/s0) is x-axes, so you can call them S or anything. i want interpret (Rmax/Ro) with regarding S.

Rmax/R0 = b * S^x.

this what I need, hope now is clear.

Thanks a lot

[Star Strider]

O.K. That much is now clear.

I still need to know what your data are. I assume in ‘num.fig’ that ‘s/s0’ is ‘x’. Is that correct? What is ‘y’? I am still not certain what the plot represents, or what the data are that are plotted.

[Ahmed Abdalazeez]

yes, in .fig

x-axes: s/s0

y=axes: Rmax/R0*(L/lambda)^1/4.

AND, (L/lambda)^1/4 is constant

[Star Strider]

How am I supposed to estimate Rmax/R0, or Rmax or R0 if that calculation has already been done?

What is the value of (L/lambda)^1/4?

[Ahmed Abdalazeez]

How am I supposed to estimate Rmax/R0?

OK, what I need, I just need a perfect fitting with these data R VS S

Regarding to the value of (L/lambda)^1/4 = (20/70)^1/4.

[Star Strider]

I am not getting the information I need.

[Ahmed Abdalazeez]

Suppose, it as experimental data distributed as in the figure above. What I have here the value of Rmax/R0 is calculated directly. So, from these data i need the best fit as power function for example Rmax/R0 = a*(s/s0)^x.

I tried the formula above but I not sure if is the best fitting or not.

[Ahmed Abdalazeez]

[Ahmed Abdalazeez]

may now it is more clear