Working Out F When You Only Have G
This usually shows up in two places: functional equations where you're given a relationship between F and G, or substitution problems in calculus where G is your u-substitution and you need to rewrite the integral entirely in G terms. The approach differs depending on which camp you're in, but both come down to the same mechanical process—eliminate every variable that isn't G. I keep seeing students mix these two up and then wonder why their answer has x in it when the question clearly said "in terms of G." Don't make that mistake.
Find F In Terms Of G
The core method is isolation and substitution. Start by identifying what relationship connects F and G. Is it an equation like F(x) = 2G(x) + x? Or is it more implicit, like F(G(x)) = x + G(x)^2? The path forward depends entirely on what you're actually given. If it's explicit, just replace every non-G variable. If G(x) = x^2 and F(x) = 3x + 5G(x), then F in terms of G is straightforward: F = 3sqrt(G) + 5G. That's it. The square root comes from inverting the G definition. You have to check whether the inverse is even valid in your domain—G(x) = x^2 only has a real inverse for x >= 0, so if your problem involves negative x values, you need G^(-1)(x) = sqrt(x) with the appropriate sign adjustment or you write it piecewise. Implicit relationships are where things get messy. Say you're told F(G(x)) = x + 3G(x) and G(x) = ln(x). You can't just solve for x algebraically in one step. What I do is substitute the known G expression first: F(ln(x)) = x + 3ln(x). Then I let u = ln(x), which means x = e^u. Now F(u) = e^u + 3u, so F in terms of the argument of G is F = e^G + 3G. The trick is recognizing that once you isolate G as your new variable, everything else becomes a function of it.
Here's a case that bit me last year. A student had F(G(x)) = G(x)^3 + 2G(x) and was asked to find F(t) in terms of t, where t = G(x). The obvious answer is F(t) = t^3 + 2t. But G(x) happened to be sin(x), which has a range of [-1, 1], so technically F is only defined on that interval. When they later tried to evaluate F(2), the whole thing fell apart because the original functional relationship never covered that input. I had them write the domain restriction next to their answer every time. It took ten seconds and prevented half the exam errors in that class. For calculus integration problems, "find F in terms of G" usually means rewriting an antiderivative. You've computed that the integral equals some expression in G, and now you substitute back. The common error here is dropping the constant or forgetting that G might itself contain a coefficient. If G = 3x + 1 and your integral gives you G^2/2, that's correct, but if you then say the answer is x^2 you've lost the scaling factor. Always check by differentiating your result in terms of G and confirming it matches the original integrand. A few things that trip people up regularly. First, G doesn't always have to be a simple single-variable function. Sometimes G is a vector or a matrix expression, and "in terms of G" means expressing F using matrix operations on G. Second, if G appears inside another function, you can't always cleanly separate it. F(G(x), H(x)) where both G and H are given might not be expressible solely in G unless H is itself a function of G. Third, inverse functions don't always exist globally. If G is not one-to-one over your domain, you'll need to restrict the domain or use a piecewise definition, and most textbook problems skip mentioning this entirely.
Get the Full Details

When the relationship between F and G involves derivatives or integrals—like F'(x) = G(x) + F(x)·G'(x)—you're dealing with a differential equation, not a simple substitution. The standard form becomes F' - GF = xG', which is linear in F. You'd use an integrating factor of exp(-G dx). This comes up more often than people expect in engineering coursework. Don't treat every appearance of F and G together as algebra. Check whether derivatives or integrals are involved before you try to isolate. The method I use for verification is substitution backward. Once you have F expressed in G, plug in a concrete value for x, compute G(x), compute F from your new expression, and confirm it matches the original equation. It's a three-second check that catches approximately 80 percent of algebra mistakes. I learned that from watching students lose points on midterms and it's saved me time ever since.