rtaylor
- obtain the Taylor series for an ODE or PDE system
Calling Sequences
rtaylor(
solved
,
options
)
rtaylor(
solved
,
vars
,
options
)
Parameters
solved -
system in solved form
vars -
(optional) solving variables of the system
options -
(optional) sequence of options to specify the ranking for the solved form, initial data, and the order of the Taylor series
Description
-
The
rtaylor
function uses an output
rifsimp
form to obtain local Taylor series expansions for all dependent variables in the ODE or PDE system simultaneously. The Taylor series output is a list containing equations of the form
depvar(indepvars)=Taylor series
.
-
The ranking related options that are accepted by
rtaylor
include the specification of the
vars
as a ranking, and the
ranking
and
indep
options described in
rifsimp[ranking]
,
rifsimp[options]
, and
rifsimp[adv_options]
.
-
Note
: specification of different
vars
than those used to obtain the result from
rifsimp
can give incomplete results.
-
The
order=n
option specifies the order that the Taylor series should be computed to, and must be a non-negative integer. The default value is
2
.
-
The
point=[ivar1=val1,...]
option specified an expansion point for the series. When this option is used, every independent variable must be given a value.
-
The table resulting from a call to
initialdata
can be given as an option to
rtaylor
, in which case the Taylor series will be given in terms of the functions present in the initial data.
-
In addition, the arbitrary functions and constants on the right hand sides of the specified initial data can be given specific values, and the expansion can be computed for these values.
-
In general, any
Constraint
or
DiffConstraint
relations (see
rifsimp[nonlinear]
) in the
rif
form cannot be used in an automatic way, so they are ignored. These relations must be accounted for manually after the Taylor series calculation. Special care must be taken when
DiffConstraint
relations are present, because
all
derivatives of these relations must be manually accounted for. This is not the case for
Constraint
relations, as they are purely algebraic.
-
The requirement that the input solved form be in
rif
form can be relaxed mildly, but
rtaylor
still requires that the equations are in a valid solved form that matches the input ranking (given in the options), and have no integrability conditions remaining. Only when these conditions hold is the resulting Taylor series an accurate representation of the local solution.
Examples
>
with(Rif):
A simple ODE
>
rtaylor([diff(f(x),x,x)=-f(x)],order=4);
A PDE system with a single dependent variable
>
rtaylor([diff(f(x,y),y,y)=diff(f(x,y),x)*f(x,y),
diff(f(x,y),x,x)=2*f(x,y)], order=3);
A PDE system with two dependent variables
>
rtaylor([diff(f(x,y),x,x)=diff(g(x,y),y),
diff(f(x,y),y,y)=diff(g(x,y),x),
diff(g(x,y),x)=diff(g(x,y),y)]);
An example using initial data
>
sys := {diff(f(x,y),x,x)=0,diff(f(x,y),x,y)=0};
>
id := initialdata(sys);
>
rtaylor(sys, id, order=3);
An example using specified initial data and an expansion point
>
ids := eval(eval(id),{_F1(y)=sin(y),_C1=1});
>
rtaylor(sys, ids, order=3, point=[x=1,y=Pi]);