The Euler (one step method) method only uses values at
to compute
We can get better estimates for
if we make use of some of the previous values.
If
then![]()
Using the Euler method gives
for step size
If we weight the values of
and
then![]()
We find
and
such that the above is exact for linear functions.
Put
respectively.
![]()
![]()
We solve the simultaneous equations
(1)
(2)
Sub (1) into (2) and
cancels to give
then from (1)![]()
Euler's equation becomes
This is a two step method.
We can also derive a three step method.
![]()
Put
respectively.
![]()
![]()
![]()
We solve the simultaneous equations
(3)
(4)
(5)
Rearrange (4) to give
and substitute (3) to give
(6)
Rearrange (5) to give
and substitute (5) and (6) to give
(7)
(7)-(6) gives
then from (6)
and finally from (3)![]()
Then we have![]()
An alternative method of finding the coefficients exists using Lagrange polynomials.
is the Lagrange polynomial such that
then![]()