How to code gaussian error function

In regard to coding, I wonder how you could execute the error function to any function, like C = atan((2*a)/(2*b)). So I'd expect Matlab to form the partial derivations of a and b and plug them into DeltaC = ((dC/da)^2 + (Delta a)^2 + (dC/db)^2 + (Delta b)^2)^(1/2). Pretending that I assigned values to a, b, Delta a, Delta b.

Answer 1

[SAA]

Matlab has a built in function for both the error function and the complimentary error function.

erf
erfc

https://www.mathworks.com/help/matlab/ref/erf.html

https://www.mathworks.com/help/matlab/ref/erfc.html

[Niklas Kurz]

Yea, I confounded that. Actually I meant error propagation. I even don't really grasp what erf is about. What are the values corresponding to? For Example if I type in erf(1) it gives back 0.8427. Do you know the meaing of it?

[SAA]

its the solution to this integral multiplied by , its used in statistics and to solve some PDE's