Theorem 5.2.1 (Substitution)
Let

and

be integrable functions and let

also be a differentiable
function. Then
 |
(5.2.1) |
with

.
Proof.
Define
 |
(5.2.2) |
or in other terms:

. So
 |
(5.2.3) |
Applying the chain rule and the fact that

gives us
 |
(5.2.4) |
Putting

and then integrating the left and the right hand side then gives
 |
(5.2.5) |
Remark 5.2.2
Formula (
5.2.1) can be more than just a bit confusing. When you use substitution,
you essentially change the variable you are integrating over. Instead of integrating
over

, you are now integrating over

, so you also have to adapt the integration
boundaries when calculating a definite integral:
 |
(5.2.6) |
Remark 5.2.4
Again, the comment has to be made that for definite integral, one should not forget
to put in the correct boundaries:
 |
(5.2.9) |
Remark 5.2.5
And this concludes pretty much all the basic techniques that you have at hand to calculate
an integral. Apart from substitution and integration by parts, there are however a lot
of standard tricks which are not really within the realm of integration theory, but
still can make your integrating life a lot easier. Here we present just a few of them:
- Look for obvious solutions. Maybe you are being asked to integrate an odd function
(a function for which
) over a symmetric interval (an interval like
). Then the picture is something like
200pt
Look e.g at the integral
 |
(5.2.10) |
for any
we choose. Remarking that the positive and the negative parts cancel
out saves us having to calculate a rather unpleasant integral.
- Use elementary algebra to rewrite the integrand in a simpler form, e.g.
 |
(5.2.11) |
- Sometimes doing a long division before integrating simplifies things a
little. The integral
 |
(5.2.12) |
looks quite unpleasant because no simplifying is possible. Rational functionals
are only easy to deal with if the degree of the numerator happens to be one
less than the degree of the denominator, and we just get a logarithm.
In this case, doing the long division of
by
gives
 |
(5.2.13) |
and therefore
- Use one of the 4 million trig formulas you were pestered with in
your more juvenile days to rewrite the integrand in a simpler form, e.g.
The most useful ones tend to be
 |
(5.2.16) |
as the following example will illustrate.
- A standard substitution technique used to integrate root functions goes
as follows:
- In
, substitute
to obtain
.
- In
, substitute
to obtain
.
- In
, substitute
to obtain
.
Example:
Using our substitution
and therefore
.
Furthermore, we know
 |
(5.2.18) |
which gives
- What is strictly speaking also part of ``elementary algebra" is the use of
partial fractions. They are very useful when having to calculate the integral
of a rational function:
 |
(5.2.20) |
Would be an easy enough integral if only there where an
in the numerator,
but now it looks kind of difficult. The standard way to tackle this problem is
to split up the fraction in different ones:
 |
(5.2.21) |
This means that
 |
(5.2.22) |
So we then have to solve the system of equations for
and
:
 |
(5.2.23) |
The solution of this system is
and
so
 |
(5.2.24) |
and therefore
There are a lot of much more complicated integrals that can be solved using this
method, but this calculation gives the general idea of it.
- When everything else fails, there is always an integral table ...