91Ó°ÊÓ

The directional derivative is a way to determine the rate of change of a function as one moves in a specific direction. It combines partial derivatives and the direction, encoded as a unit vector.
Think of it as answering the question: "If I move in this precise direction, how fast does my function change?"
To find the directional derivative of a function \( f \) at a point \( (x, y) \), use a unit vector \( \vec{u} = (a, b) \). The formula is:

• \( f_{\vec{u}}(x, y) = f_x(x, y) \cdot a + f_y(x, y) \cdot b \)

In this formula:- \( f_x \) and \( f_y \) are the partial derivatives, indicating the function's rate of change in the x and y directions.
- \( a \) and \( b \) are components of the unit vector specifying direction.The critical part is ensuring the unit vector follows \( a^2 + b^2 = 1 \). Only then is the direction correctly normalized.

Partial derivatives \( f_x \) and \( f_y \) are a cornerstone of calculus, representing how a multivariable function changes with respect to each individual variable while keeping all other variables constant.
They answer the simple question: "How does the function change if I nudge one variable while holding the others fixed?"
To compute partial derivatives:

• For \( f_x \), differentiate the function concerning \( x \) while treating \( y \) as a constant.
• For \( f_y \), differentiate the function concerning \( y \) while treating \( x \) as a constant.

In every problem involving rate of change, partial derivatives lay the groundwork. You can picture them as the building blocks of more complex derivative concepts like directional derivatives and gradients.

The gradient vector is created by combining all the partial derivatives of a function into a single vector, \( abla f = \langle f_x, f_y \rangle \).
It's like a compass that points in the direction of the steepest increase of a function.
When you look at a hill on a terrain map, the gradient shows which path is steepest uphill. This makes it a potent tool in optimization, helping us find maxima and minima.
In the context of directional derivatives, the gradient vector gets paired with a unit vector. This pairing gives you the directional derivative through the dot product:

• Directional Derivative = \( abla f \cdot \vec{u} \)

Because the dot product involves multiplication, decreasing the directional derivative's value requires choosing \( \vec{u} \) such that the angle between \( abla f \) and \( \vec{u} \) is more than 90 degrees. This chooses the vector that's "against the grain," resulting in \( f_{\vec{u}} < 0 \).