Next: 8.3 A Non-integrable Function
 Up: 8. Integrable Functions
 Previous: 8.1 Definition of the
     Index 
  8.9   
Definition (Operations on functions.) 
Let 

 and 

 be functions where 

 are sets. 
Let

.  We define functions 

, 

, 

, 

 and 

 as
follows:
 
Remark:  These operations of addition, subtraction, multiplication
and division for functions satisfy the associative, commutative and
distributive laws that you expect them to.  The proofs are straightforward and
will
be omitted.
  8.10   
Definition (Partition-sample sequence.) 
Let 
![$[a,b]$](img1071.gif)
 be an interval.  By a 
partition-sample sequence for 
![$[a,b]$](img1071.gif)
 I
will mean a pair of sequences 

 where 

 is a sequence
of
partitions of 
![$[a,b]$](img1071.gif)
 such that
 

, and for each 

 in 

,

 is a sample for 

.
 
  8.11   
Theorem (Sum theorem for integrable functions.) 
Let 
 be integrable functions on an interval 
.  Then 
 and
are integrable on 
 and
and
 
Proof:  Suppose 
 and 
 are integrable on 
.  Let 
be
a partition-sample sequence for 
.  If 
 and
, then
Since 
 and 
 are integrable, we have
By the sum theorem for sequences,
Hence 
 is integrable and 
. 
The proof of the second statement is left as an exercise.
  8.12   
Notation (
) 
If 

 is integrable on an interval 
![$[a,b]$](img1071.gif)
 we will sometimes write

 instead of 

.  The ``

" in this expression is
a dummy variable, but the ``

" is a part of the notation and may not be
replaced by another symbol.  This notation will be used mainly in cases where
no particular name is available for 

.  Thus
means
 

 where 

 is the function on 
![$[1,2]$](img1125.gif)
 defined by

 for all 
![$t\in [1,2]$](img2042.gif)
.  The ``

" here stands for difference,
and 

 is a ghost of the differences 

 that appear in the
approximations
for the integral.  The 

 notation is due to Leibniz.
 
  8.13   
Example. 
Let 
This function is integrable over every closed bounded subinterval of 

,
since it is a sum of five functions that are known to be
integrable. By several applications of the sum theorem for
integrals we get
 
  8.15   
Theorem (Inequality theorem for integrals.) Let 
 and 
be integrable functions on the interval 
 such that
Then
 
  8.16   
Exercise. 
Prove  the inequality theorem for integrals.
A
 
Proof: We have
Hence by the inequality theorem for integrals
Hence
It follows that
  8.18   
Theorem.
Let 
 be real numbers with 
, and let 
 be a function
from 
 to 
.  Suppose 
 is integrable on 
 and 
 is
integrable on
.  Then 
 is integrable on 
 and 
. 
Proof:
Since 
 is integrable on 
 and on 
, it follows that
 is bounded on 
 and on 
, and hence 
 is bounded
on 
.
Let 
 be a partition-sample sequence for 
. 
For each 
 in 
 we define a partition 
 of 
 and a
partition
 of 
, and a sample 
 for 
,
 and a
sample
 for 
 as follows:
Then there is an index 
 such that 
. 
 
Let
We have
where
Let 
 be a bound for 
 on 
.  Then
Also,
Now
Since
it follows from the squeezing rule that 
and hence 
.
From equation (8.21)
we have
  | 
(8.22) | 
 
Since 
 and 
, we
see
that 
 is a partition-sample sequence on
, and 
 is a partition-sample
sequence on
.  Since 
 was given to be integrable on 
 and on 
, we
know
that
and
Hence it follows from (8.22) that
i.e., 
 is integrable on 
 and
  8.24   
Definition (Spike function.) 
Let 
![$[a,b]$](img1071.gif)
 be an interval. A function 
![$f:[a,b] \to \mbox{{\bf R}}$](img2103.gif)
 is  called
a 
spike function, if there exist numbers 

 and 

, with
![$c \in [a,b]$](img2104.gif)
 such that
 
Proof:  Case 1: Suppose 
 Observe that 
 is increasing on the interval 
 and decreasing
on the interval 
, so 
 is integrable on each of these intervals.
The set of points under the graph of 
 is the union of a horizontal
segment and a vertical segment, and thus is a zero-area set. Hence
By the previous theorem, 
 is integrable on 
, and
.
Case 2: Suppose 
. Then by case 1 we see that 
 is integrable with
integral equal to zero,
so by the sum theorem for integrals 
 too. 
  8.28   
Exercise. 
Prove corollary 
8.26, i.e., explain why it follows from theorem
8.25.
 
  8.29   
Definition (Piecewise monotonic function.) 
A function 

 from an interval 
![$[a,b]$](img1071.gif)
 to 

 is 
piecewise monotonic
if there are points 

 in 
![$[a,b]$](img1071.gif)
 with 

such that 

 is monotonic on each of the intervals 
![$[a,a_1],[a_1,a_2],\cdots
,[a_{n-1},a_n],[a_n,b]$](img2126.gif)
.
 
  8.30   
Example. 
The function whose graph is sketched below is piecewise monotonic.
 
  8.31   
Theorem.
Every piecewise monotonic function is integrable.
Proof:  This follows from corollary 8.23. 
  8.32   
Exercise. 
A
Let
Sketch the graph of 

.  Carefully explain why 

 is integrable, and find

.
 
  8.33   
Example. 
Let 

 Then
Hence 

 is integrable on 
![$[0,3]$](img2133.gif)
, and
 
 
 
  
 
 Next: 8.3 A Non-integrable Function
 Up: 8. Integrable Functions
 Previous: 8.1 Definition of the
     Index 
Ray Mayer
2007-09-07