91Ó°ÊÓ

The gradient of a function is an essential concept in multivariable calculus. It is like a compass showing you the direction of the steepest ascent of a function. For a function of several variables, the gradient is represented as a vector. This vector comprises the partial derivatives of the function.
The gradient gives us important insight:

• Direction of greatest increase of the function
• Magnitude reflects the rate of increase

To find the gradient of a function like \( \phi(x, y, z) = x^2 y z + 4x^2 \), you calculate the vector of partial derivatives: \( abla \phi(x, y, z) = \left( \frac{\partial \phi}{\partial x}, \frac{\partial \phi}{\partial y}, \frac{\partial \phi}{\partial z} \right) \). This means taking the derivative of the function concerning each variable while treating other variables as constant.

In our example, these partial derivatives provide:

• \( \frac{\partial \phi}{\partial x} = 2xy z + 8x \)
• \( \frac{\partial \phi}{\partial y} = x^2 z \)
• \( \frac{\partial \phi}{\partial z} = x^2 y \)

So, the gradient \( abla \phi \) becomes a vector \( (2xy z + 8x, x^2 z, x^2 y) \). Evaluating this vector at a point, such as \( (1, -2, -1) \), reveals the direction in which the function increases most rapidly from this specific point in space.

Partial derivatives are a fundamental component in understanding changes in multivariable functions. A partial derivative of a function at a point gives the rate at which the function changes as one of the input variables changes, with all other variables held constant.

This is pivotal because, unlike single-variable functions that vary along one axis, multivariable functions can change along multiple axes, giving rise to a partial derivative for each variable.

• Consider \( \phi(x, y, z) = x^2 y z + 4x^2 \). Here, to find the partial derivative with respect to \( x \), treat \( y \) and \( z \) as constants: \( \frac{\partial \phi}{\partial x} = 2xy z + 8x \).
• For \( y \), treat \( x \) and \( z \) as constants: \( \frac{\partial \phi}{\partial y} = x^2 z \).
• And for \( z \), keep \( x \) and \( y \) constant: \( \frac{\partial \phi}{\partial z} = x^2 y \).

This helps isolate the effect each variable has on the function. It also lays the foundation for constructing the gradient vector, which consolidates these partial derivatives to analyze function behavior collectively from any given point in space.

In the context of directional derivatives, a unit vector plays a crucial role. A vector becomes a unit vector when it has a magnitude of exactly one. It is used to indicate direction without scaling effects of different magnitudes. A unit vector helps ensure we're only considering the direction, not the length of the vector.

To convert a vector to a unit vector:

• Find its magnitude. For example, the vector \( \overrightarrow{PQ} = (-2, 0, 4) \) has a magnitude of \( \sqrt{(-2)^2 + 0^2 + 4^2} = \sqrt{20} \).
• Divide each component of the vector by this magnitude, resulting in \( \hat{u} = \left(-\frac{1}{\sqrt{5}}, 0, \frac{2}{\sqrt{5}}\right) \).

When computing directional derivatives, you're essentially projecting how fast the function increases in the given unit direction. The unit vector ensures this direction remains consistent, offering a standardized way to measure directional derivatives independent of the actual size of the original direction vector.