Percentage agreement between data sets - automate

I am currently trying to automate several processes that I do repeatedly. I have recently gotten the stats toolbox. Is there a function on there that automatically calculated percentage agreement between theoretical and experimental datasets? For example X = DATASET (A large vector of over two hundred values) and Y = Theoretical data

Answer 1

[the cyclist]

Something like this may work for you:

X = [1.000 2.004 3.05];
Y = [1 2 3];

tolerance = 0.01;

percentAgreement = 100*sum((abs(X-Y) < tolerance))/numel(A)

Answer 2

[the cyclist]

I don't think there will be anything "out of the box" that will do exactly what you want, but this seems very easy to program to me. You'll need to be precise in what you mean by "agreement", and address some simple logistical issues.

For example

• Are you doing pairwise comparison, or population-level comparison?
• Are the theoretical and experimental datasets already in one-to-one correspondence, or do you need to do some sorting?
• Does agreement need to be perfect, or is 1.00001 equal to 1?

etc.

[Bran]

Ahh thought not. I am doing pairwise and I have already reorganised the datasets. No not perfect so yes in this case 1.00001 would be acceptable