Error Analysis of the Euler Method
As before, we are considering the first-order initial value problem
and approximating its solution using Euler's method with a certain
step size . I will ignore roundoff error and consider only the
discretization error.
For
step-by-step methods such as Euler's for solving ODE's, we want to
distinguish between two types of discretization
error: the global error and the local error.
The global error at a certain value of (assumed to be
)
is just what we would ordinarily call the error: the difference between
the true value and the approximation .
The local error at is, roughly speaking, the error introduced in
the th step of the process. That is, it is the difference between
and
, where is the solution
of the differential equation with
. Thus if
were exactly correct (equal to ), the global
error at would be equal to this local error. But since in
general is not correct (as a result of earlier local errors),
the global and local errors are different.
In the picture below, is the black curve,
and the curves are in red.
The local errors at each stage of the process
are the blue vertical lines.
First we consider the local error at :
Now
. According to Taylor's Theorem, for any
twice-differentiable function
for some between and . Taking
, and we find
If there is some constant such that we can be sure that
, then we can say
Such a does exist (assuming has continuous derivatives in some
rectangle containing the true and approximate solutions):
for any solution of the differential equation , we can differentiate
once more to get
which is a continuous function of and and therefore
can't get too big in our rectangle.
Now, what about the global error? It's tempting to say that the
global error at is the sum of all the local errors for
from 1 to . Since each and there are
of them, the global error should be
.
Unfortunately, it's not quite true that the global error is the sum of the
local errors: the global error at is the sum of the differences
, but
.
Between and ,
might grow or shrink.
Fortunately, we can control the amount of growing that might take
place, and the result is that it grows by at most some constant factor
(again, this is in a rectangle where has continuous derivatives and
which contains the true and approximate solutions).
Let's look at a simple example: , . This is so simple
that we can find an explicit formula for . We have
Thus at each stage is multiplied by . Since we start out with
, we have
The actual solution is of course
. The
global error at with step size , where , is
Since
the local error in step is
.
The local error is because (from Taylor series)
. The global error is : in fact,
on the
O and Order page, we used the example
, which we saw had error . The
Euler approximation is just , so it too has error .
Another special case: suppose is just a function of .
The true solution is
For the Euler method we have
, so that
This is a just a Riemann sum for the integral: for each interval
we are approximating the area under the graph of
by a rectangle with height . As you may have seen in
calculus, the error in this approximation for each interval is
at most if on the interval, and the global
error at is then at most
.
Robert
2002-01-28