Compute the directional derivative of at , in the direction of .
We’ve defined the directional derivatives of a function, which allow us to determine how a function is changing in various directions.
However, we would like an easier way to evaluate directional derivatives, that doesn’t require the limit definition.
Such a method will require use of the gradient of the function. Recall our definition of the gradient of a function .
We’ll see that this vector turns out to be closely related to directional derivatives.
The gradient and directional derivatives
Let’s suppose we have a differentiable function , and consider our definition of the directional derivative a function at in the direction of a unit vector , which was We’ll rewrite this definition by considering another function, . Notice that is a single variable function, and when we rewrite the directional derivative, we have
So, the directional derivative is just the derivative of this single variable function evaluated at . Revisiting our definition of , we can use the chain rule to find the derivative of .
Evaluating at , we have
Thus, we have arrived at the following result.
Let’s use this result to compute some directional derivatives.
Since isn’t a unit vector, we need to normalize it. Since , we’ll use the vector to compute our desired directional derivative.
Next, we’ll need the gradient of . Since the partial derivatives of are polynomials, they are continuous, so is differentiable. Thus, we can use the above theorem to compute the directional derivative.
Then, we can compute the directional derivative as
This matches what we had previously computed using the definition of directional derivatives.
The gradient and level sets
We’ve shown that for a differentiable function , we can compute directional derivatives as What does this mean for the possible values for a directional derivative? Recall that the dot product can be computed as where is the angle between the two vectors. Since is a unit vector, we have Since , we have that In particular, the largest that can be is , and this occurs when points in the same direction as , so that . Thus, the gradient points in the direction of greatest increase.
On the other hand, the minimum value that can have is , and this occurs when points in the opposite direction from , in the direction of . Thus, points in the direction of greatest decrease.
Additionally, from , we can see that is perpendicular to if and only if . But what does it mean for ? This means that there is no instantaneous change in in the direction of , which means that will be a tangent vector to a level curve.
Let’s consider how this function changes near the point .
The gradient of at is Consulting the graph of near the point , we can confirm that it increases most rapidly when we move in the direction of . We can also confirm that it decreases most rapidly when we move in the direction of .
The point lies on the level curve . A tangent vector to this level curve at the point is given by .
This vector, , is perpendicular to our gradient .
We’ll state this observation more formally, and prove that the gradient is perpendicular to the level curves.
- Proof
- We need to show that for any vector which is tangent to at , we have
that is perpendicular to .
If is tangent to , we can find a parametrized curve lying in such that and . We will show that is perpendicular to .
By the definition of , and since lies in , for all . Differentiating both sides of this identity, and using the chain rule on the left side, we obtain Plugging in , this gives us which we can rewrite as Thus, we have shown that is perpendicular to the level set .