Recognizing Separable Equations

Recognizing the Product Structure

A first-order differential equation describes a relationship between an independent variable \(x\), a dependent variable \(y\), and the derivative \(y'=\frac{dy}{dx}\). In the notation established in the earlier material on first-order differential equation notation, an equation can be written as \(y'=F(x,y)\), where \(F\) gives the slope at each point in its domain.

A useful feature to look for is whether this slope can be expressed as a product: one factor depending only on \(x\), and another depending only on \(y\). Recognizing that structure identifies a separable equation. At this stage, the goal is only to recognize and, when needed, rearrange the equation into that form. We will not solve it here.

The factors need not be complicated. For example, \(x^2\) is an \(x\)-only factor, while \(1+y^2\) is a \(y\)-only factor. By contrast, a sum such as \(x+y\) does not display a product of an \(x\)-only factor and a \(y\)-only factor. A sum might sometimes be rewritten by factoring, but it is not separable merely because it contains both variables.

Directly Separable Equations

In a directly separable equation, the derivative is already expressed as a product with the required dependence. Recognizing the factors is a matter of checking that each factor contains only its indicated variable. The same expression can have many equivalent factorizations; the important property is that at least one valid factorization exists on the domain under consideration.

It is not necessary for both factors to be nonconstant. For instance, an equation \(y'=3x\) is also a product-form equation: take \(f(x)=3x\) and \(g(y)=1\). This observation is useful, but it does not make every expression separable. The dependence on the variables still has to factor into a product.

Rearranging into Product Form

Some equations place \(y'\) alongside a coefficient that depends on \(x\), \(y\), or both. Algebra may isolate \(y'\) and expose a product. When doing this, keep track of any value that must be excluded in order to divide. A rearranged equation is equivalent to the original only where the algebraic operation used is valid.

The restriction matters even when the final expression looks familiar. For example, if the coefficient of \(y'\) is \(x+1\), dividing by it excludes \(x=-1\). That exclusion is a restriction on the points where the rearranged equation represents the original one; it should not be silently discarded.

A Test for Whether a Product Factorization Is Possible

When the right-hand side is not obviously a product, a simple identity can help test it. Suppose \(F(x,y)=f(x)g(y)\). For any two \(x\)-values \(x_1,x_2\) and two \(y\)-values \(y_1,y_2\) in the domain, multiplying the opposite corners gives equal results:

This equality is a necessary condition for a product factorization. On a rectangular domain, it is also sufficient when \(F\) is nowhere zero. The nonzero condition lets us define a factor using division by a fixed value of \(F\).

Here is the proof. If \(F=f(x)g(y)\), the equality follows by multiplying the factors in either order, as shown above. For the converse, choose fixed \(x_0\in I\) and \(y_0\in J\), and define

The denominator is nonzero because \(F\) is nowhere zero. Apply the assumed identity with \(x_1=x\), \(x_2=x_0\), \(y_1=y\), and \(y_2=y_0\). It gives

Dividing by \(F(x_0,y_0)\), which is nonzero, yields

This proves both directions. If \(F\) has zeros, the identity remains necessary, but this particular converse requires extra care; the nonzero condition is part of the theorem.

Recognizing a Nonseparable Right-Hand Side

The test can rule out a product form by finding just one choice of four points where the identity fails. The following calculation is a concrete way to distinguish a sum that mixes the variables from a separable product.

This test should be applied on the domain in question. If the equation is defined only on a smaller region, the four selected points must lie in that region. Failure at valid points proves that a product factorization is impossible there. Passing a few numerical checks, however, does not establish factorability; the identity in the theorem must hold for all relevant choices.

Common Pitfall: Dividing Without Checking

A frequent mistake is to divide by a factor that might be zero, then treat the resulting equation as equivalent everywhere. For instance, from \((y-3)y'=2x+1\), division by \(y-3\) creates an expression defined only when \(y\ne 3\). Writing the rearranged equation as if it also applied at \(y=3\) is not valid: the denominator is zero there. The correction is to state the restriction and remember that the rearrangement represents the original equation only on that restricted domain.

A related mistake is to call an equation separable just because its right-hand side contains an \(x\)-term and a \(y\)-term. The expression \(x+y\) has both variables, but the opposite-corner test shows that it does not factor as required. Separation is about a product of single-variable factors, not merely about variables appearing in the same formula.

A Recognition Checklist

When you are given a first-order equation, use the following process before attempting any further method:

Recognizing this structure matters because it tells you which method may be appropriate next, before doing any solving. Just as importantly, noticing a domain restriction prevents a rearrangement from being mistaken for a universally equivalent equation. The key question is precise: can the slope be expressed as an \(x\)-only factor times a \(y\)-only factor, and on what domain is that expression valid?