Pure · Integration
1 / 10
Integration by substitution
The formal version of the reverse chain rule — substitute u = g(x), replace dx using du = g'(x) dx, integrate in u, then convert back (or change the limits for a definite integral).
Pure · Integration
Integration by substitution
The formal version of the reverse chain rule — substitute u = g(x), replace dx using du = g'(x) dx, integrate in u, then convert back (or change the limits for a definite integral).
Why it works
Integration by recognition works when you can see the answer. Substitution is the systematic version for when you can't: you rename the awkward inner part as a new variable , and the whole integral simplifies into something standard.The mechanics rest on one substitution for the differential. If , then You treat like a fraction here (legitimate inside an integral): wherever appears, replace it by , and replace the inner by . Every must disappear — the new integral has to be entirely in — or the substitution isn't the right one.
Take with . Then , so becomes and becomes : Often you'll need to rearrange the substitution to replace a stray too — e.g. from you also get to deal with an extra factor of .
Definite integrals: change the limits. With a definite integral you have two clean choices — and the slick one is to convert the -limits into -limits using , integrate in , and evaluate. Then you never convert back. For with : when , ; when , ; so it becomes .