Find R square with different length vectors

Hello!

I have two vectors, first with 50 elements and second with 3000 elements. Second vector is a Gompertz Function modeling of the first, and I need to calculate the R2, but in all the functions that I found the problem is I need two same size vectors... Can you help me?

Answer 1

[Image Analyst]

You can't as far as I know. You can only correlate the 50 "test" data values with the 50 corresponding modeled versions of them -- those same values run through the model. How did you even get the other 250 input values to do the model on?

[Vinicio Moya]

With fit function, but, with those data I generate a ANN, and now I need to compare de R2 into the RAW and ANN points.

[Image Analyst]

If you make up additional fitted values, you can't correlate them. For example if I have an (x1, y1) and (x2, y2) and get a line between them, and now I get 1000 points in between x1 and x2 and get fitted y values, I can't correlation those additional 1000 y values to anything because they don't exist. Only y1 and y2 exist. So you could correlate yFit1 and yFit2 to y1initial and y2initial, but not y(2) - y(999) of the fitted data because there is no corresponding y values to correlate against.

[Vinicio Moya]

ok, I get it... Thanks!!