91Ó°ÊÓ

When we talk about the gradient in mathematics, we're dealing with a vector that represents the slope of a function in multiple dimensions. Think of it as a compass that points in the direction of the steepest climb of a function. For a function like \(h(x, y, z)\), the gradient \(abla h\) is calculated using the partial derivatives of the function with respect to each variable (x, y, z).

• The partial derivative with respect to \(x\) shows how \(h\) changes as \(x\) changes, while keeping \(y\) and \(z\) constant.
• The same logic applies for the derivatives with respect to \(y\) and \(z\).

The gradient is assembled from these partial derivatives, giving us a vector that points in the direction where the function increases most rapidly. Each component of this vector tells us how steep the slope is along each axis. With our original function, the gradient at a point \(P_0\) is \(abla h(P_0) = (1, \frac{1}{2}, 2)\), representing the slope in the x, y, and z directions respectively.

The concept of partial derivatives is foundational when analyzing functions of several variables. A partial derivative measures how a multivariable function changes as one of its variables is altered, while all other variables are kept fixed.
For instance, if we consider the function \(h(x, y, z) = \cos(xy) + e^{yz} + \ln(zx)\), each partial derivative calculates the rate of change along one specific axis:

• \(h_x = -y\sin(xy) + \frac{1}{x}\)
• \(h_y = -x\sin(xy) + ze^{yz}\)
• \(h_z = ye^{yz} + \frac{1}{z}\)

These derivatives help create the gradient vector, each component representing the function's sensitivity to changes in each variable. They tell us how curved or steep the surface represented by the function is along each particular dimension.

Normalization is a process used to transform a vector into a unit vector—meaning it has a length of one. This is particularly useful when we're interested in the direction of a vector rather than its magnitude. Say we have a vector \(\mathbf{u} = \mathbf{i} + 2\mathbf{j} + 2\mathbf{k}\).

• First, calculate its magnitude: \(\|\mathbf{u}\| = \sqrt{1^2 + 2^2 + 2^2} = 3\).
• Then, divide each component by the magnitude to get the normalized vector: \(\mathbf{u}_{normal} = (\frac{1}{3}, \frac{2}{3}, \frac{2}{3})\).

This normalized vector now maintains the same direction but has a magnitude of 1, making it ideal for directional calculations like the directional derivative. Using a unit vector simplifies many problems and allows us to focus on direction purely.

The dot product is a key operation in vector algebra, offering a way to multiply two vectors. Given two vectors \(\mathbf{a} = (a_1, a_2, a_3)\) and \(\mathbf{b} = (b_1, b_2, b_3)\), their dot product is calculated as:\[\mathbf{a} \cdot \mathbf{b} = a_1 b_1 + a_2 b_2 + a_3 b_3\]This operation results in a scalar, providing information about the degree to which two vectors point in the same direction. In the context of directional derivatives, the dot product between the gradient and a normalized direction vector tells us how fast the function is changing in that specific direction.
For example, using \(abla h(P_0) = (1, \frac{1}{2}, 2)\) and \(\mathbf{u}_{normal} = (\frac{1}{3}, \frac{2}{3}, \frac{2}{3})\), their dot product gives us the directional derivative, calculated as:\[1 \cdot \frac{1}{3} + \frac{1}{2} \cdot \frac{2}{3} + 2 \cdot \frac{2}{3} = 2\]This value represents the rate of change of the function \(h\) at point \(P_0\) in the direction of \(\mathbf{u}\).