Is possible solve a integral with a vector in the limitis?

%I am tring to make a fit where the xdata are in the limits of integral

%but this is not allowed

clear all

load C:\Users\Aluno\Desktop\rafael\novo\B100.txt

xdata = B100_(:,1);

ydata = B100_(:,2);

fun = @(x,xdata) x(1) + x(2).*integral( @(V) V , 0 , xdata);

x0 = [ 1, 2];

output = lsqcurvefit(fun,x0,xdata,ydata)

[Jan]

What is the mathematical meaning of an integral with a vector as limit? As long as it is not defined mathematically, you cannot expect that a standard function calculates it.

[eric rocha]

For example, I am trying make a fit with the result of a implicit integral that only can be solved numericaly. (sorry about my bad gramma, I am a portuquese native speak and my english is not very good)

Answer 1

[Star Strider]

Yes.

Try this:

fun = @(x,xdata)  x(1) + x(2).*arrayfun(@(xd)integral( @(V)  V , 0 , xd),xdata);

It works either with or without the 'ArrayValued',1 name-value pair. The arrayfun function replaces a for loop, so each element of ‘xdata’ is evaluated as the upper limit of the integration in each iteration.

[eric rocha]

Really thanks.

[Star Strider]

As always, my pleasure.

Answer 2

[Walter Roberson]

In order to evaluate integral with multiple limits you will need to use arrayfun.

If you are concerned about efficiency you could sort xdata and integrate between consecutive endpoints and cumsum and then permute back to the original order .