To much points for mathcad 2000???

[RoyRobRobson]

I want to calculate a graph in very small stepps. It seems it is to much for mathcad because I allways get the error that it is not possible to plot so many points. Is there anywhere a way to change the amount of possible points?

Cheers

 

[IRstuff]

How many points did you need? TTFN

 

[RoyRobRobson]

Well, as more points as better is the calculation. At the moment I think mathcad allows abaut 300000-500000 points but I want to go 2 till 4 times higher.

 

[IRstuff]

Since there is nothing that I know of that would allow you to resolve an individual data point on such a plot, perhaps you're using plot to solve some problem completely unrelated to plotting? TTFN

 

[RoyRobRobson]

Well, I have a complicate function with some very sharp and steep peaks. The only interesting thing is the value right at the top of the peaks. If I set the steps of calculation to big, maybe I get one value of the peak but at the next step the calculation is allready behind the peak. Mathcad is not able to deviate this function so I can´t calculate the position of the peaks to get only the value of these points.

 

[IRstuff]

Seems to me that there are two alternate solutions:

1) animate the plot with the desired step size tied to the animation frame number. This would result in a sequence of plot slices that you can view with adequate magnification and speed.

2) write a program to peak find TTFN

 

[GregLocock]

I'd use MathCAD to sort out the peaks.

Something like (in pseudo code)
j=0
for i = 1 to (numpoints in data array)-1
if data(i)>(i-1)and data(i)>data(i+1) then
j=j+1
peaks(j)=data(i)
endif
next i

To do this in an older version of mathcad may need a bit of finagling but I'm sure it is possible since it is the kind of thing I used to do all the time.

Hey, as a general point can people please identify which version they are using when they ask a question?



Cheers

Greg Locock

 

[GregLocock]

I've figured it out for the early versions, go to

http://www.geocities.com/greglocock/index.html

hit gallery, and it is most of the way down.

This is a general method, using an array of indirection pointers.

Many apparently iterative problems are solvable in early versions of mathcad using arrays of pointers, but they are not intuitive, to this little black duck at least.

Cheers

Greg Locock

 

[RoyRobRobson]

Well, I have done it with program and it works fine. At the moment I get the position and the value of the peak in a field. Now the next problem comes up.. I want the hole function depending on parameter. The amount of peaks, the value and the position would depend on this parameter.
But k(x)_i(x) is not possible. "_" is standing for index. How can I go around this problem? Its the first time I´m using mathcad, so one problem apears after the other... With a matrice the amount of peaks could only go up to 100 as far as I know.

Thanks for helping me! Rob

 

[GregLocock]

Could you post your file somewhere, and an explanation of the problem?
What is your definition of a hole function?

k[x,i[x should be legitimate, I think, although as written i can see you'd need to generate a 2d array of pointers.

Cheers

Greg Locock

 

[RoyRobRobson]

For example:

d:=2
a(x):=x^2
i:=1..a(d)
k_i:=d/2

k_i=
1
1
1
1

This is ok in mathcad, but if I want k and i depending on d it is not possible.


a(x):=x^2
i(d):=1..a(d) this is not allowed
k(d)_i(d):=0.5d and so is this also not ok.

Maybe I could live with this version:

a(x):=x^2
i:=1..3
k(d)_i:=0.5d

but why is it not ok??? Any way to fix it???

 

[GregLocock]

i(d):=1..a(d) this is not allowed

Because you are trying to set each individual element of i to a range. Each element of i can only be a number.

"k(d)_i(d):=0.5d and so is this also not ok. "

Don't know why that doesn't work if i(d) is just a number.


Cheers

Greg Locock