problems with fitting curve from matrix data

Hey there, i have some problems to fitting a curve. Both matrix have two columns and the same length, but i get the error:

??? XDATA must be a matrix with one to two columns.

Error in ==> fit at 115 errstr = handleerr('curvefit:fit:xDataMustBeColumnVector', ...

Error in ==> Niederschlag_Trend at 127 [fitobject, R2] = fit(x,y, 'linearinterp');

What I need is the determination coefficient for the double mass curve, if there is any suggestion to solve it another way, i´m happy to hear it :)

    %%Homogenitätstest
    load(['niederschlag2_' num2str(NS.start_jahr) '-'num2str(NS.end_jahr)]);
    load ref81-2010.mat;
    testdata = MaxMatrix(:,2);
    refdata = RefMat(:,2);

    x=testdata(1);
    y=refdata(1);
    for i=2:length(testdata);
        x(i)=x(i-1) + refdata(i);
        y(i)=x(i-1) + testdata(i);
    end
    %%xi=[min(x),max(x)];
    %%yi=[min(y),max(y)];

    figure(3)
    clf
    plot(x,y,'*k','LineWidth',2, 'Color','black')
    hold on

    *[fitobject, R2] = fit(x,y, 'linearinterp');*  
    %%plot(xi,yi,'r','LineWidth',2,'Color','red');

Answer 1

[Titus Edelhofer]

Hi Eleen,

hard to guess but probably x and y are row vectors. I would preallocate to be sure:

x = zeros(size(testdata));
y = zeros(size(testdata));
x(1) = testdata(1);
y(1) = refdata(1);
for i=2:...

This way x and y are column vectors which fit will better like. BTW, take a look at the function cumsum, which might replace the loop).

Titus