Quotient Rules with Variable-Dependent Denominators

Why Variable-Dependent Denominators Need Care

A rational function of several variables is a quotient whose numerator and denominator may each change as the variables change. When finding a partial derivative, both changes matter whenever both depend on the variable of differentiation.

The main task is to apply the quotient rule with the other variables held fixed, then simplify the result while the variables are still symbolic. Substituting a particular point before completing this work can hide terms or make it harder to see whether the denominator’s derivative has been included.

Key idea: For a quotient \(f=N/D\), the partial derivative with respect to a variable includes both the partial derivative of \(N\) and the partial derivative of \(D\). The formula is valid where \(D\ne 0\), provided the relevant partial derivatives exist.

The Partial Quotient Rule

Let \(N\) and \(D\) be differentiable functions of \(x\) and \(y\), and suppose \(D(x,y)\ne 0\). For \(f(x,y)=N(x,y)/D(x,y)\), the partial derivatives are

f_x(x,y) = \frac{D(x,y)N_x(x,y)-N(x,y)D_x(x,y)} {D(x,y)^2}, \qquad f_y(x,y) = \frac{D(x,y)N_y(x,y)-N(x,y)D_y(x,y)} {D(x,y)^2}.

Here, \(N_x\) and \(D_x\) describe changes with respect to \(x\), holding \(y\) fixed; \(N_y\) and \(D_y\) describe changes with respect to \(y\), holding \(x\) fixed. The same rule applies to any variable: differentiate the numerator and denominator with respect to that variable, and keep the other variables fixed.

The order of the terms in the numerator matters: it is denominator times derivative of numerator, minus numerator times derivative of denominator. The denominator of the resulting fraction is the square of the original denominator.

Why the Rule Works

The quotient rule follows from the product rule and the derivative of a reciprocal. To see this for \(x\), write \(f=N D^{-1}\). Differentiating the identity \(D D^{-1}=1\) with respect to \(x\) gives \(D_xD^{-1}+D(D^{-1})_x=0\). Because \(D\ne 0\), this implies \((D^{-1})_x=-D_x/D^2\). Now apply the product rule to \(N D^{-1}\):

f_x = N_xD^{-1}+N(D^{-1})_x = \frac{N_x}{D}-\frac{ND_x}{D^2} = \frac{DN_x-ND_x}{D^2}.

The calculation for \(f_y\) is identical with \(x\) replaced by \(y\). This also explains why omitting the denominator’s derivative is an error: the reciprocal factor \(D^{-1}\) changes whenever \(D\) changes.

A Reliable Calculation Process

For a quotient with several variable-dependent parts, writing down the ingredients before combining them helps prevent sign and substitution errors.

Name the numerator and denominator.
Write \(f=N/D\), and note that the calculation is restricted to points where \(D\ne 0\).

Compute the needed partials.
Find \(N_x,D_x\) for \(f_x\), or \(N_y,D_y\) for \(f_y\), holding the other variable fixed.

Apply the quotient rule.
Use \(DN_x-ND_x\) in the numerator and \(D^2\) in the denominator.

Simplify before evaluating.
Combine like terms and retain the variables. Substitute a point only if an evaluation is requested.

Worked Examples

Worked Example: A Linear Denominator in Two Variables

Find both first partial derivatives of \(f(x,y)=(x^2+3y)/(x-y)\), and state the domain condition. Set \(N=x^2+3y\) and \(D=x-y\). The quotient is defined when \(x-y\ne 0\).

For the \(x\)-partial, \(N_x=2x\) and \(D_x=1\), since \(y\) is held fixed. Thus

\begin{aligned} f_x &=\frac{(x-y)(2x)-(x^2+3y)(1)}{(x-y)^2}\\ &=\frac{2x^2-2xy-x^2-3y}{(x-y)^2}\\ &=\frac{x^2-2xy-3y}{(x-y)^2}. \end{aligned}

For the \(y\)-partial, \(N_y=3\) and \(D_y=-1\). Therefore

\begin{aligned} f_y &=\frac{(x-y)(3)-(x^2+3y)(-1)}{(x-y)^2}\\ &=\frac{3x-3y+x^2+3y}{(x-y)^2}\\ &=\frac{x^2+3x}{(x-y)^2}. \end{aligned}

As a check, take the point \((2,1)\), which is in the domain. There \(N=7\), \(D=1\), and the formulas give \(f_x=-3\) and \(f_y=10\). The differentiated identity \(N_x=D_xf+Df_x\) gives \(4=7-3\), and \(N_y=D_yf+Df_y\) gives \(3=-7+10\). Both checks agree.

Worked Example: A Quadratic Denominator

Find \(f_x\) and \(f_y\) for \(f(x,y)=(2x+y^2)/(x^2+y+1)\). Let \(N=2x+y^2\) and \(D=x^2+y+1\). The domain consists of points where \(x^2+y+1\ne 0\). The needed partials are \(N_x=2\), \(D_x=2x\), \(N_y=2y\), and \(D_y=1\).

Applying the rule with respect to \(x\) gives

\begin{aligned} f_x &=\frac{2(x^2+y+1)-(2x+y^2)(2x)}{(x^2+y+1)^2}\\ &=\frac{2x^2+2y+2-4x^2-2xy^2}{(x^2+y+1)^2}\\ &=\frac{-2x^2-2xy^2+2y+2}{(x^2+y+1)^2}. \end{aligned}

With respect to \(y\), the numerator and denominator both have nonzero partial derivatives:

\begin{aligned} f_y &=\frac{(2y)(x^2+y+1)-(2x+y^2)(1)}{(x^2+y+1)^2}\\ &=\frac{2x^2y+2y^2+2y-2x-y^2}{(x^2+y+1)^2}\\ &=\frac{2x^2y+y^2+2y-2x}{(x^2+y+1)^2}. \end{aligned}

Check at \((1,1)\): \(N=3\), \(D=3\), \(f=1\), and the formulas give \(f_x=0\), \(f_y=1/3\). For \(x\), \(N_x=2=D_xf+Df_x=2(1)+3(0)\). For \(y\), \(N_y=2=D_yf+Df_y=1+3(1/3)\). The symbolic derivatives pass both checks.

Worked Example: Both Variables in a Shared Quadratic Denominator

Find the partial derivatives of \(f(x,y)=(xy+1)/(x^2+y^2)\). Here \(N=xy+1\) and \(D=x^2+y^2\). The denominator is zero only at \((0,0)\), so the quotient and these formulas apply wherever \(x^2+y^2\ne 0\). The required partials are \(N_x=y\), \(D_x=2x\), \(N_y=x\), and \(D_y=2y\).

For \(x\), substitute those quantities and simplify:

\begin{aligned} f_x &=\frac{y(x^2+y^2)-(xy+1)(2x)}{(x^2+y^2)^2}\\ &=\frac{x^2y+y^3-2x^2y-2x}{(x^2+y^2)^2}\\ &=\frac{y^3-x^2y-2x}{(x^2+y^2)^2}. \end{aligned}

For \(y\), the denominator derivative is \(2y\), not zero:

\begin{aligned} f_y &=\frac{x(x^2+y^2)-(xy+1)(2y)}{(x^2+y^2)^2}\\ &=\frac{x^3+xy^2-2xy^2-2y}{(x^2+y^2)^2}\\ &=\frac{x^3-xy^2-2y}{(x^2+y^2)^2}. \end{aligned}

At the valid point \((1,1)\), \(N=2\), \(D=2\), and \(f=1\). Both displayed partial derivatives equal \(-1/2\). For \(x\), \(N_x=1=D_xf+Df_x=2-1\); for \(y\), \(N_y=1=D_yf+Df_y=2-1\). This confirms the signs and the denominator contributions at that point.

A Common Mistake: Treating the Denominator as Constant

In the first example, an incorrect calculation of \(f_x\) might differentiate only the numerator and write \(2x/(x-y)\). That expression treats \(x-y\) as though it were constant with respect to \(x\), even though its partial derivative is \(1\). For instance, at \((2,1)\) this incorrect expression gives \(4\), whereas the correct partial derivative is \(-3\).

The missing term is the contribution from differentiating the denominator. The quotient rule gives \(((x-y)(2x)-(x^2+3y)(1))/(x-y)^2\), which simplifies to \((x^2-2xy-3y)/(x-y)^2\). At \((2,1)\), this is \(-3\), as in the symbolic calculation. A useful check before simplifying is to ask whether the denominator depends on the variable being differentiated; if it does, its partial derivative belongs in the quotient-rule numerator.

Why Symbolic Simplification Comes First

A partial derivative is a function of the variables, defined on the part of the original domain where its formula makes sense. Leaving the variables visible preserves that information and makes the result useful at any permitted point. Substituting a point too early can produce a number without showing the derivative function, and may conceal a missing term or an invalid denominator.

In each example, the denominator of the derivative is the square of the original denominator. Therefore the same restriction \(D\ne 0\) remains essential. Do not cancel or discard this restriction merely because a simplified numerator happens to vanish at some point. The original quotient is still undefined wherever its original denominator is zero.

Takeaway: Identify \(N\) and \(D\), compute the partials with the other variables fixed, apply \((DN_x-ND_x)/D^2\) or its corresponding-variable version, and simplify before substituting any point.

Check Your Understanding

For each question, keep the variables symbolic unless the question explicitly asks for a value.

For \(f(x,y)=(x+y)/(x-2y)\), what domain condition is required, and what are \(N_x,D_x,N_y,D_y\)?
Apply the partial quotient rule to find \(f_x\) for \(f(x,y)=(x^2+y)/(x+y^2)\). State where the formula applies.
For \(g(x,y)=(xy)/(x^2+1)\), which term in the quotient-rule numerator vanishes when finding \(g_y\), and why?
A learner claims that the \(x\)-partial of \((x+y)/(x+y+1)\) is \(1/(x+y+1)\). Identify the omitted contribution and write the corrected quotient-rule expression.
For a quotient \(N/D\), explain how the identity \(N_x=D_xf+Df_x\) can be used to check a proposed formula for \(f_x\).