Multivariable calculus

Gradient of x^2 y z^3: Monomial in Three Variables

For f(x, y, z) = x^2 y z^3 the gradient is (2x y z^3, x^2 z^3, 3x^2 y z^2). In each component only the exponent belonging to that variable changes: it comes down as a factor and drops by one, while the other two powers ride along untouched. At (2, 3, 1) this is (12, 4, 36).

f(x,y,z)=x2yz3f(x,y,z) = x^2 y z^3
f=2xyz3,  x2z3,  3x2yz2\nabla f = \left\langle 2xyz^3,\; x^2 z^3,\; 3x^2 y z^2 \right\rangle
fx=2xyz3\frac{\partial f}{\partial x} = 2xyz^3
fy=x2z3\frac{\partial f}{\partial y} = x^2 z^3
fz=3x2yz2\frac{\partial f}{\partial z} = 3x^2 y z^2

The power rule, applied to one variable at a time

A monomial makes the gradient mechanical. To differentiate with respect to xx, treat yz3yz^3 as a constant multiplier and apply the power rule to x2x^2.

x(x2yz3)=yz32x=2xyz3\frac{\partial}{\partial x}\left(x^2 y z^3\right) = yz^3 \cdot 2x = 2xyz^3

For yy the exponent is 1, so it comes down as a 1 and yy disappears entirely. For zz the exponent 3 comes down and leaves z2z^2.

fy=x2z3,fz=3x2yz2\frac{\partial f}{\partial y} = x^2 z^3, \qquad \frac{\partial f}{\partial z} = 3x^2 y z^2

There is a shortcut worth knowing for monomials. Away from the coordinate planes you can factor ff out of every component, since fx=f2x\frac{\partial f}{\partial x} = f \cdot \frac{2}{x} and similarly for the others.

f=x2yz32x,  1y,  3z\nabla f = x^2 y z^3 \left\langle \frac{2}{x},\; \frac{1}{y},\; \frac{3}{z} \right\rangle

The mistake: lowering every exponent at once

The most common wrong answer is fx=2xz2\frac{\partial f}{\partial x} = 2xz^2 or 6xyz26xyz^2, produced by differentiating two variables in the same step. A partial derivative changes exactly one exponent.

  • Check the total degree. The function has degree 2+1+3=62 + 1 + 3 = 6, so every component of the gradient must have degree 5.
  • 2xyz32xyz^3 has degree 1+1+3=51+1+3 = 5, x2z3x^2z^3 has degree 2+3=52+3=5, and 3x2yz23x^2yz^2 has degree 2+1+2=52+1+2=5. All three pass.
  • A wrong answer like 6xyz26xyz^2 has degree 4, which flags the error before you substitute a single number.

The second slip is losing the yy component. Since yy appears to the first power, fy=x2z3\frac{\partial f}{\partial y} = x^2 z^3 contains no yy at all, and students sometimes read that as a mistake and put the yy back.

A numerical check at a point

At (2,3,1)(2, 3, 1) the function value is f=431=12f = 4 \cdot 3 \cdot 1 = 12.

f(2,3,1)=2231,  41,  3431=12,  4,  36\nabla f(2,3,1) = \left\langle 2 \cdot 2 \cdot 3 \cdot 1,\; 4 \cdot 1,\; 3 \cdot 4 \cdot 3 \cdot 1 \right\rangle = \left\langle 12,\; 4,\; 36 \right\rangle

The zz component is by far the largest, which the shortcut formula predicts: the exponent on zz is the biggest, and at z=1z=1 the factor 3z\frac{3}{z} beats 2x=1\frac{2}{x} = 1 and 1y=13\frac{1}{y} = \frac{1}{3}. In relative terms the exponent is the whole story: raising zz by one percent raises ff by about three percent, raising xx by one percent raises it by two, and raising yy by one percent raises it by one.

Every component carries a factor of xx and a factor of zz, so f=0\nabla f = \mathbf{0} exactly on the plane x=0x = 0 and on the plane z=0z = 0. The critical set is two whole planes. Setting y=0y = 0 alone is not enough, since fy=x2z3\frac{\partial f}{\partial y} = x^2 z^3 survives it.

Frequently asked questions

Why does yy vanish from fy\frac{\partial f}{\partial y}?

Because yy appears to the first power. Differentiating y1y^1 gives 1y0=11 \cdot y^0 = 1, so the yy factor is replaced by 1 and only x2z3x^2 z^3 remains. Nothing was lost.

Which variable changes ff fastest at (2,3,1)(2,3,1)?

The variable zz. The components are 12,4,36\left\langle 12, 4, 36 \right\rangle, so a small step in zz changes ff three times as fast as the same step in xx and nine times as fast as one in yy.