I can't connect to that link.
Take logs: .
Differentiate implicitly: .
We have: .
I read the theory over and over again, but i don't get it.
You want the derivative of x^x. The theory says
"If , then . The chain rule is then used, to differentiate ln(y) with respect to x (y'/y). The product rule is then used on the right. Then solve for y'"
So you'd get
ln (y)'--> (1/y) =ln(x)+1<-- xln(x)'
Thus y= 1/ln(x)+1, which is wrong. The derivative of x^x should be:x^x(ln(x)+1)
1) Where is the mistake?
2)as the graph of y=f(x) is the same as the graph of ln(y)=ln(f(x)), why can you not simply derive the "right side", why do you have to derive both sides? In "normal" deriviation, you also don't derive the y!
The other issue is the "rule of logarithmic differentiation" which says that .
This follows from differentiation of with respect to x by the chain rule: . Solving for gives the above rule.
And, yes, differentiation of can be done by referring to this rule of logarithmic differentation (if you happen to know it):