eval giving the wrong mathematical answer

I'm trying to write a script that I will use for Newton Rhapson approximation. I am testing as I go and I noticed that the eval command is returning the incorrect mathematical answer. I am testing it with x=2 with func=((2*x)*(log(x)))-1

Here is the code:

syms x;
func=input('Enter your function in terms of x: ');
f=@(x)(func);
z=diff(f(x));
f1=@(x)(z);

x0=input('Enter your initial value of x: ');
x=x0
s=char(f1);

y=subs(f1)

eval(y)

Upon entering x=2 and func=((2*x)*(log(x)))-1

The code redturns:

x =

       2

y =

2*log(2) + 2

ans =

      3.3863

The answer to 2*log(2)+2 should be 2.602

What am I doing wrong?

Answer 1

[Roger Stafford]

In matlab the 'log' function is considered to be the natural logarithm with base e, not 10. Therefore 3.3863 is the correct answer, not 2.602 .

[John D'Errico]

And of course, if you really wanted a log to the base 10, use log10.

2*log10(2)+2 
ans =
     2.6021

Computers are so silly. They do what you tell them to do, instead of what you want them to do.