Multivariable calculus
Partial Derivatives of (x + y)/(1 + z^2)
For f(x,y,z) = (x + y)/(1 + z^2), the partial derivatives with respect to x and with respect to y are both 1/(1 + z^2), and the partial with respect to z is negative 2z(x + y) divided by (1 + z^2) squared. Only the z partial needs the quotient rule.
Ask which variable is actually in the denominator
A fraction does not automatically call for the quotient rule. What matters is whether the differentiating variable appears in the denominator. Here and appear only upstairs, so for those two the whole factor is a constant multiplier.
That makes the first two partial derivatives almost immediate. Differentiating with respect to gives , and the constant factor comes along.
For the situation is different, since sits in the denominator. Treat as a constant and differentiate . The quotient rule gives numerator over .
The mistake: reaching for the quotient rule every time
Running the quotient rule for is not wrong, it is just slower and it invites algebra errors. It gives , which simplifies back to . Students who do not simplify often leave an answer that looks different from the correct one and then distrust it.
The genuine error to avoid is a sign. Differentiating a function that sits in a denominator produces a minus sign, and it is easy to lose when you are also juggling the constant . Rewriting as makes the sign automatic: the power rule turns into and leaves a factor of out front.
Reading the answers back
At the denominator is , so both the and partials equal . The partial is .
The signs are worth interpreting. Increasing or increases the numerator, so those partials are positive everywhere. The partial carries the factor , so when it is negative for and positive for : moving away from zero in either direction grows the denominator and shrinks the fraction. On the plane the partial vanishes, since the denominator is momentarily flat.
Frequently asked questions
Why do the x and y partial derivatives come out equal?
Because the function depends on and only through the sum , and that sum changes at the same rate whichever of the two you move. Any function of has equal partial derivatives in and .
Is this function defined everywhere?
Yes. The denominator is at least for every real , so it never vanishes and no point has to be excluded. That is what makes the quotient safe to differentiate anywhere.