# Explain proof of directional derivative

I will try to mark the exact thing I don't understand, it comes from this thread https://matchmaticians.com/questions/u6me4r#answers :
What I don't understand is when we have $$g(h)=f(x_0+ha,y_0+hb)$$
That means that when 'h'(independent variable) changes in g(h), it will produce the same change in 'h' of f(x0​+ha,y0​+hb).
Derivative formula says that:
$$g'(x) = \lim_{{h \to 0}}\frac{g(x+h)-g(x)}{h}$$
so just changing variable name I receive
$$g'(h) = \lim_{{\Delta{h} \to 0}}\frac{g(h+\Delta{h})-g(h)}{\Delta{h}}$$
And when I want to find it's derivative at 0 I receive:
$$g'(0) = \lim_{{\Delta{h} \to 0}}\frac{g(0+\Delta{h})-g(0)}{\Delta{h}} = \lim_{{\Delta{h} \to 0}}\frac{g(\Delta{h})-g(0)}{\Delta{h}}$$
So the question is since $$g(h)=f(x_0+ha,y_0+hb)$$
Why can I replace $$g(\Delta{h}) by f(x_0+ha,y_0+hb)$$
Since by replacing it I will receive not $$f(x_0+ha,y_0+hb)$$ but $$f(x_0+\Delta{h}a,y_0+\Delta{h}b)$$
P.S. I understand how directional derivatives work, I don't want you to explain it, I just want the explanation of this particular case please, feel free to ask for more money,.

Answers can be viewed only if
1. The questioner was satisfied and accepted the answer, or
2. The answer was disputed, but the judge evaluated it as 100% correct.
• Erdos
0

Let me know if you need any clarifications.

• Babaduras
0

yeah, I think I was confusing delta(h) and (h) all the time. so correct case for derivative of g at 0 is g'(0) = limit of (g(delta(h) - g(0))/delta(h) with delta(h) approaching zero, replaceing g(h) with f(x0​+ha,y0​+hb) will give me limit of (f(x0​+delta(h)a,y0​+(delta(h)b) - f(x0​,y0​))/delta(h) with delta(h) approaching zero ?

• Erdos
0

Replacing g(h) with f(x0​+ha,y0​+hb) will give me limit of (f(x0​+ha,y0​+hb) - f(x0​,y0​))/h with h approaching zero. You statement "replaceing g(h) with f(x0​+ha,y0​+hb) will give me limit of (f(x0​+delta(h)a,y0​+(delta(h)b) - f(x0​,y0​))/delta(h) with delta(h) approaching zero" has the same problem. Remember, you either use h or \Delta h. It does not make sense to use both in one computation.

• Babaduras
0

yeah, I got it, I mean that since g(h) = f(x0​+ha,y0​+hb), then replacing h with delta(h) will give me imit of (f(x0​+delta(h)a,y0​+(delta(h)b) - f(x0​,y0​))/delta(h) with delta(h) approaching zero

• Erdos
0

That's correct.