Main Index
Mathematical Analysis
Real Analysis
Subject Index
comment on the page
Two essential restrictions are made when developing the Riemann’s integral , both
should be bounded. In many applications these restrictions can be avoided. When either the integrand or the integration domain (or both) are unbounded the resulting integral is called improper. Thus we get two basic types of improper integrals:
Let and function be unbounded as approaches (from the left). The improper integral is said to converge if the limit exists and is finite. If this holds we say that the value of the limit is the value of the improper integral. If the limit above does not exist or is improper we say that the improper integral diverges. An analogous definition can be formulated when the integrand is unbounded at the lower limit of the integration. Similarly, when the integrand has improper one-sided limits at an inner point of interval but the integrand is defined on , then
provided both improper integrals on the right-hand side exist.
Often it is difficult to find the corresponding antiderivative which is necessary to answer the divergence of convergence of the given improper integral using the definition. As the terminology convergent or divergent indicates there are close connections between the theory of improper integrals and infinite series. As it was the case with infinite series also here the problem of convergence is easier to answer as the problem of finding the values to which the given improper integral converges. Various comparison arguments are of great help.
Comparison test: Let both functions and are unbounded at and for all . Then
The improper integral is said to converge if there exists the finite limit . In this case we also say that the value of the limit is the value of the improper integral. If the limit above does not exist or is improper we say that the corresponding improper integral diverges.
If the improper integral is split into a sum of improper integrals (because presents more than one improper behavior on ), then the integral converges if and only if any single improper integral is convergent.
Comparison test: Assume that for all .Then
Comparison with function of the type with a real exponent gives the following
-test: the improper integral
Limit test: Let and be two positive function defined on . Assume that both functions exhibit an improper behavior at and
Then is convergent if and only if is convergent.
The above tests of convergence for improper integrals are only valid for positive functions. Since there is a very natural way of generating a positive function from a given function just take its absolute value, consider a function (not necessarily positive) defined on and the positive function still defined on . Both functions and its absolute value exhibit the same kind of improper behavior over . Therefore, one may ask naturally what conclusion can be made if we know something about the integral
We have the following partial answer:
If the integral is convergent, then the integral is also convergent.
This shows that the convergence of carries more information than just convergence. Therefore we say that if the improper integral converges that is absolutely convergent. And if the improper integral is convergent while the improper integral is divergent, we say it is conditionally convergent.
There is a parallelism between improper integrals and infinite series. The integral test bridges the two notions. One of its form states (for more details consult ) :
Integral test. Let be a function defined for , bounded, positive and monotonically decreasing to 0 as . Let
Then for the difference tends to a finite limit.
In particular, the series is convergent if and only if the integral exists and is finite.
Cite this web-page as:
Štefan Porubský: Improper Integral.