Do dlgradient and dlfeval have to be used together?

Do dlgradient and dlfeval have to be used together?

function [y,dydx] = rosenbrock(x)
y = 100*(x(2) - x(1).^2).^2 + (1 - x(1)).^2;
dydx = dlgradient(y,x);
end

[y,dydx] = rosenbrock([1,2]),

get error??? Value to differentiate must be a traced dlarray scalar.

Can't it be as convenient as pytorch？”a traced dlarray scalar.“ not clear!!!

Answer 1

[KALYAN ACHARJYA]

x=[1,2];
[y,dydx]=rosenbrock(x);

[cui,xingxing]

Undefined function 'dlgradient' for input arguments of type 'double'.

should type "dlarray"

Answer 2

[cui,xingxing]

but i input like this in command window , is OK ,why???

[fval,gradval] = dlfeval(@rosenbrock,dlarray([1,2]))

fval =

1×1 dlarray

100

gradval =

1×2 dlarray

-400 200

[hang Jiang]

try:

x=dlarray(x);

dlarray(y)=100*(x(2)-x(1).^2).^2+(1-x(1)).^2;