Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I want to show that if $f:\mathbb{R}^n\to \mathbb{R}$ and $df_a$ is the derivative of the function at $a$ then $df_a(v)=\displaystyle\frac{\partial f}{\partial v}(a)$. I saw a few proofs of this proposition, but all of them were pretty much the same (using the same definition for derivability) and I wanted do a proof using another definition.

I have a sketch, but I believe it may have some troubles. Instead of proving that $df_a(a)=\displaystyle\frac{\partial f}{\partial v}(a)$ I show that $df_a(tv)=\displaystyle\frac{\partial f}{\partial tv}(a)$ considering the limit

$$\lim_{tv\to 0}\frac{|f(a+tv)-f(a)-\frac{\partial f}{\partial tv}(a)|}{|tv|}\\=\lim_{t\to 0}\frac{|f(a+tv)-f(a)-t\frac{\partial f}{\partial v}(a)|}{|t||v|} \\\leq\lim_{t\to 0}\frac{|f(a+tv)-f(a)-t\frac{\partial f}{\partial v}(a)|}{|t|}\\=\lim_{t\to 0}\left|\frac{f(a+tv)-f(a)-t\frac{\partial f}{\partial v}(a)}{t}\right|\\=\left|\lim_{t\to 0}\frac{f(a+tv)-f(a)}{t}-\frac{\partial f}{\partial v}(a)\right|=0.$$

Follows that $df_a(tv)=\displaystyle\frac{\partial f}{\partial tv}$ and since $df_a(tv)=tdf_a(v)$ and $\displaystyle\frac{\partial f}{\partial tv} = t \displaystyle\frac{\partial f}{\partial v}$ is $df_a(v)=\displaystyle\frac{\partial f}{\partial v}$. Now I'm unsure about the second step, whether is legitimate to change $tv\to 0$ by $t\to 0$, which would make no differencefor $f(a+tv)$ but I have some doubts if this remains true for the partial derivative as well.

share|improve this question
Is $f$ a function of one variable? –  5xum Jul 16 at 14:06
@5xum No, $f:\mathbb{R}^n\to \mathbb{R}$ –  Cure Jul 16 at 15:59

1 Answer 1

up vote 0 down vote accepted

I suppose you want to show the directional derivative can be defined as follows: $$ \frac{\partial}{\partial v}f(x)=\lim_{t\rightarrow 0}\frac{f(x+tv)-f(x)}{t} $$ where $v$ is usually taken to be a unit vector. The above limit can be computed via some textbook tricks like $$ f(x+x',y+y')-f(x,y+y')+f(x,y+y')-f(x,y)\approx x'\frac{\partial}{\partial x}f(x,y+y')+y'\frac{\partial}{\partial y}f(x,y)+o(x'^{2}+y'^{2}) $$for the two dimensional case. For the general case it is similar. For this to work we need to assume that all the first order derivatives are continuous, so when the vector $(x',y')$ is small enough the above approximates $x'\frac{\partial}{\partial x}f(x,y)+y'\frac{\partial}{\partial y}f(x,y)$ as desired. I think there must be better ways to do this, though.

share|improve this answer

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.