5D-Interpolation of scattered Data

Hello everyone,

i have the following problem. Several tests were made over certain operating points (Xn). As a result, the measured value (Y) was recorded and stored. Unfortunately, not all operating points could be approached. Therefore, some lines in Y are empty.

As an example, I've attached a simple Excel-File with invented (but logical) values.

My goal is to interpolate between the discrete operating points (Xn).

For Example, my matlab-function should determinate a Y-value for (I think, linear-interpolation should be ok):

X1 = 2.2

X2 = 28

X3 = 180

X4 = 3

Does anyone have an idea how to do that?

Thank you very much!

I am very excited!

Answer 1

[Rishabh Singh]

If you want to calculate the value "Y", for any certain combination of { X1, X2, X3,X4 } you can look into linear regression. Also as your data has "NaN" values, you should use remmissing to remove missing values out from your data. Also consider looking into feval for evaluation of function value.

data = readtable("Example.xlsx");
X = cell(4,1);
for i =1:4
    X{i} = data(:,i);
end

data = table2array(rmmissing(data));
Y = data(:,5);
mdl = fitlm(data(:,1:4), Y)
output = feval(mdl, 2.2,28,180,3);

The above code is just a simple implementation, you can look into the linear regression documentation for additional information.

Hope this helps.