Namespaces
Variants
Actions

Difference between revisions of "Butcher series"

From Encyclopedia of Mathematics
Jump to: navigation, search
(Importing text file)
 
 
(9 intermediate revisions by 3 users not shown)
Line 1: Line 1:
''B-series''
+
<!--
 +
b1110701.png
 +
$#A+1 = 106 n = 0
 +
$#C+1 = 106 : ~/encyclopedia/old_files/data/B111/B.1101070 Butcher series,
 +
Automatically converted into TeX, above some diagnostics.
 +
Please remove this comment and the {{TEX|auto}} line below,
 +
if TeX found to be correct.
 +
-->
  
Let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110701.png" /> be an [[Analytic function|analytic function]] satisfying an ordinary differential equation <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110702.png" />. A Butcher series for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110703.png" /> is a series of the form
+
{{TEX|auto}}
 +
{{TEX|done}}
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110704.png" /></td> </tr></table>
+
''B-series''
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110705.png" /> is the set of rooted trees, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110706.png" /> is the coefficient for the series associated with the tree <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110707.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110708.png" /> is the elementary differential associated with the tree <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b1110709.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107010.png" /> is a stepsize or <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107011.png" />-increment, and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107012.png" />, often called the order of the tree <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107013.png" />, is the number of nodes in the tree <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107014.png" />. All terms will be defined below. For a more thorough discussion of Butcher series, see [[#References|[a1]]] or [[#References|[a2]]].
+
Let  $  y : \mathbf R \rightarrow {\mathbf R  ^ {d} } $
 +
be an [[Analytic function|analytic function]] satisfying an ordinary differential equation  $  y  ^  \prime  ( x ) = f ( y ( x ) ) $.  
 +
A Butcher series for $  y $
 +
is a series of the form
  
==Definitions.==
+
$$
A tree is a connected acyclic graph (cf. also [[Graph|Graph]]). A rooted tree is a tree with a distinguished node, called the root (which is drawn at the bottom of the sample trees in the Table). Two rooted trees <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107015.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107016.png" /> are isomorphic (i.e., equivalent) if and only if there is a [[Bijection|bijection]] <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107017.png" /> from the nodes of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107018.png" /> to the nodes of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107019.png" /> such that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107020.png" /> maps the root of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107021.png" /> to the root of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107022.png" /> and any two nodes <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107023.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107024.png" /> are connected by an edge in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107025.png" /> if and only if <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107026.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107027.png" /> are connected by an edge in <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107028.png" />.
+
B ( a,y ( x ) ) = \sum _ {t \in T } a ( t ) F ( t ) ( y ( x ) ) {
 +
\frac{h ^ {\rho ( t ) } }{\rho ( t ) ! }
 +
} ,
 +
$$
  
The following bracket notation for rooted trees is very useful. Let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107029.png" /> and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107030.png" /> be the trees with zero and one node, respectively. For rooted trees <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107031.png" /> with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107032.png" /> for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107033.png" />, let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107034.png" /> be the rooted tree formed by connecting a new root to the roots of each of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107035.png" />. Clearly, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107036.png" /> and any rooted tree formed by permuting the subtrees <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107037.png" /> of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107038.png" /> is isomorphic to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107039.png" />. The rooted trees with at most four nodes are displayed in the table below.''''''<table border="0" cellpadding="0" cellspacing="0" style="background-color:black;"> <tr><td> <table border="0" cellspacing="1" cellpadding="4" style="background-color:black;"> <tbody> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107040.png" /></td> <td colname="2" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107041.png" /></td> </tr> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107042.png" /></td> <td colname="2" style="background-color:white;" colspan="1">
+
where  $  T $
 +
is the set of [[rooted tree]]s,  $  a ( t ) \in \mathbf R $
 +
is the coefficient for the series associated with the tree  $  t $,  
 +
$  F ( t ) ( y ( x ) ) \in \mathbf R  ^ {d} $
 +
is the elementary differential associated with the tree $  t $,
 +
$  h \in \mathbf R $
 +
is a stepsize or  $  x $-
 +
increment, and  $  \rho ( t ) $,
 +
often called the order of the tree  $  t $,  
 +
is the number of nodes in the tree  $  t $.
 +
All terms will be defined below. For a more thorough discussion of Butcher series, see [[#References|[a1]]] or [[#References|[a2]]].
  
<img style="border:1px solid;" src="https://www.encyclopediaofmath.org/legacyimages/common_img/" />
+
==Definitions==
 +
A tree is a connected acyclic graph (cf. also [[Graph|Graph]]). A rooted tree is a tree with a distinguished node, called the root (which is drawn at the bottom of the sample trees in the Table). Two rooted trees $t$
 +
and  $\widehat{t}$
 +
are isomorphic (''i.e.'' equivalent) if and only if there is a [[Bijection|bijection]] $\sigma $
 +
from the nodes of  $t$ to the nodes of  $\widehat{t}$
 +
such that  $  \sigma $ maps the root of  $  t $
 +
to the root of  $  {\widehat{t}  } $
 +
and any two nodes  $  n _ {1} $
 +
and  $  n _ {2} $ are connected by an edge in  $  t $
 +
if and only if  $  \sigma ( n _ {1} ) $
 +
and  $  \sigma ( n _ {2} ) $
 +
are connected by an edge in  $  {\widehat{t}  } $.
  
</td> </tr> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107043.png" /></td> <td colname="2" style="background-color:white;" colspan="1">
+
The following bracket notation for rooted trees is very useful. Let $\bullet$
 +
be the tree with one node. For rooted trees  $ t_{1} \dots t _ {m} $
 +
with  $  \rho ( t _ {i} ) \geq  1 $
 +
for  $  i = 1 \dots m $,
 +
let  $  t = [ t _ {1} \dots t _ {m} ] $
 +
be the rooted tree formed by connecting a new root to the roots of each of  $  t _ {1} \dots t _ {m} $.  
 +
Clearly,  $  \rho ( t ) = 1 + \sum _ {i = 1 }  ^ {m} \rho ( t _ {i} ) $
 +
and any rooted tree formed by permuting the subtrees  $  t _ {1} \dots t _ {m} $
 +
of  $  t $
 +
is isomorphic to  $  t $.
  
<img style="border:1px solid;" src="https://www.encyclopediaofmath.org/legacyimages/common_img/" />
+
The rooted trees with at most four nodes are displayed in the table below.
  
</td> </tr> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107044.png" /></td> <td colname="2" style="background-color:white;" colspan="1">
+
{| class="wikitable"
 +
|-
 +
| $t_1$ || $t_2=[t_1]$ || $ t_{3} = [t_{1},t_{1}]$ || $t_4 = [t_2]$
 +
|-
 +
| [[File:Rooted tree 0.svg|50px]] || [[File:Rooted tree 10.svg|50px]] || [[File:Rooted tree 200.svg|100px]]|| [[File:Rooted tree 110.svg|50px]]
 +
|-
 +
|}
  
<img style="border:1px solid;" src="https://www.encyclopediaofmath.org/legacyimages/common_img/" />
+
{| class="wikitable"
 +
|-
 +
| $t_5 =[t_1,t_1,t_1]$ || $t_6=[t_2,t_1]$ || $ t_{7} = [t_{3}]$ || $  t_{8} = [t_4]$
 +
|-
 +
| [[File:Rooted tree 3000.svg|100px]] || [[File:Rooted tree 2100.svg|100px]] || [[File:Rooted tree 1200.svg|100px]]|| [[File:Rooted tree 1110.svg|50px]]
 +
|-
 +
|}
  
</td> </tr> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107045.png" /></td> <td colname="2" style="background-color:white;" colspan="1">
+
Using this bracket notation for rooted trees, one can recursively define elementary differentials, as follows. Let
 
 
<img style="border:1px solid;" src="https://www.encyclopediaofmath.org/legacyimages/common_img/" />
 
 
 
</td> </tr> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107046.png" /></td> <td colname="2" style="background-color:white;" colspan="1">
 
 
 
<img style="border:1px solid;" src="https://www.encyclopediaofmath.org/legacyimages/common_img/" />
 
  
</td> </tr> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107047.png" /></td> <td colname="2" style="background-color:white;" colspan="1">
+
$$
 +
F ( \emptyset ) ( y ( x ) ) = y ( x ) , \quad F ( \bullet ) ( y ( x ) ) = f ( y ( x ) ) ,
 +
$$
  
<img style="border:1px solid;" src="https://www.encyclopediaofmath.org/legacyimages/common_img/" />
+
and, for  $  t = [ t _ {1} \dots t _ {m} ] $
 +
with  $  \rho ( t _ {i} ) \geq  1 $
 +
for  $  i = 1 \dots m $,
  
</td> </tr> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107048.png" /></td> <td colname="2" style="background-color:white;" colspan="1">
+
$$
 +
F ( t ) ( y ( x ) ) =
 +
$$
  
<img style="border:1px solid;" src="https://www.encyclopediaofmath.org/legacyimages/common_img/" />
+
$$
 +
=  
 +
f ^ {( m ) } ( y ( x ) ) ( F ( t _ {1} ) ( y ( x ) ) \dots F ( t _ {m} ) ( y ( x ) ) ) ,
 +
$$
  
</td> </tr> <tr> <td colname="1" style="background-color:white;" colspan="1"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107049.png" /></td> <td colname="2" style="background-color:white;" colspan="1">
+
where  $  f ^ {( m ) } ( y ( x ) ) $,
 +
the  $  m $
 +
th derivative of  $  f $
 +
with respect to  $  y $,
 +
is a multi-linear mapping.
  
<img style="border:1px solid;" src="https://www.encyclopediaofmath.org/legacyimages/common_img/" />
+
Since  $  y ( x ) $
 
+
satisfies  $  y  ^  \prime  ( x ) = f ( y ( x ) ) $,
</td> </tr> </tbody> </table>
+
it is not hard to show that
 
 
</td></tr> </table>
 
 
 
Using this bracket notation for rooted trees, one can recursively define elementary differentials, as follows. Let
 
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107050.png" /></td> </tr></table>
+
$$
 +
y ^ {( i ) } ( x ) = \sum _ {t \in T _ {i} } \alpha ( t ) F ( t ) ( y ( x ) ) ,
 +
$$
  
and, for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107051.png" /> with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107052.png" /> for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107053.png" />,
+
where  $  y ^ {( i ) } ( x ) $
 +
is the  $  i $
 +
th derivative of  $  y $
 +
with respect to  $  x $,  
 +
$  T _ {i} = \{ {t \in T } : {\rho ( t ) = i } \} $,
 +
and  $  \alpha ( t ) $
 +
is the number of distinct ways of labeling the nodes of  $  t $
 +
with the integers  $  \{ 1 \dots \rho ( t ) \} $
 +
such that the labels increase as you follow any path from the root to a leaf of  $  t $.  
 +
Consequently, one can write the [[Taylor series|Taylor series]] for $  y ( x ) $
 +
as a Butcher series:
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107054.png" /></td> </tr></table>
+
\begin{equation}
 +
\label{eq:1}
 +
y ( x + h ) = \sum _ {i = 0 } ^  \infty  y ^ {( i ) } ( x ) {
 +
\frac{h  ^ {i} }{i! }
 +
} =
 +
\end{equation}
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107055.png" /></td> </tr></table>
+
$$
 +
=  
 +
\sum _ {t \in T } \alpha ( t ) F ( t ) ( y ( x ) ) {
 +
\frac{h ^ {\rho ( t ) } }{\rho ( t ) ! }
 +
} .
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107056.png" />, the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107057.png" />th derivative of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107058.png" /> with respect to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107059.png" />, is a multi-linear mapping.
+
The importance of Butcher series stems from their use to derive and to analyze numerical methods for differential equations. For example, consider the  $  s $-
 +
stage Runge–Kutta formula (cf. [[Runge–Kutta method|Runge–Kutta method]])
  
Since <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107060.png" /> satisfies <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107061.png" />, it is not hard to show that
+
\begin{equation}\label{eq:2}
 +
y _ {n + 1 }  = y _ {n} + h \sum _ {i = 1 } ^ { s }  b _ {i} k _ {i} ,
 +
\end{equation}
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107062.png" /></td> </tr></table>
+
$$
 +
k _ {i} = f \left ( y _ {n} + h \sum _ {j = 1 } ^ { s }  a _ {ij }  k _ {j} \right ) ,
 +
$$
  
where <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107063.png" /> is the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107064.png" />th derivative of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107065.png" /> with respect to <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107066.png" />, <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107067.png" />, and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107068.png" /> is the number of distinct ways of labeling the nodes of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107069.png" /> with the integers <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107070.png" /> such that the labels increase as you follow any path from the root to a leaf of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107071.png" />. Consequently, one can write the [[Taylor series|Taylor series]] for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107072.png" /> as a Butcher series:
+
for the ordinary differential equation  $  y  ^  \prime  ( x ) = f ( y ( x ) ) $.  
 +
Let  $  b = ( b _ {1} \dots b _ {s} )  ^ {T} \in \mathbf R  ^ {s} $
 +
be the vector of weights and  $  A = [ a _ {ij }  ] \in \mathbf R ^ {s \times s } $
 +
the coefficient matrix for (a2). Then it can be shown that $  y _ {n + 1 }  $
 +
can be written as the Butcher series
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107073.png" /></td> <td valign="top" style="width:5%;text-align:right;">(a1)</td></tr></table>
+
\begin{equation}\label{eq:3}
 +
y _ {n + 1 }  = \sum _ {t \in T } \alpha ( t ) \psi ( t ) F ( t ) ( y _ {n} ) {
 +
\frac{h ^ {\rho ( t ) } }{\rho ( t ) ! }
 +
} ,
 +
\end{equation}
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107074.png" /></td> </tr></table>
+
where the coefficients  $  \psi(t) \in \mathbf R $,
 +
defined below, depend only on the rooted tree  $t$
 +
and the coefficients of the formula \eqref{eq:2}.
  
The importance of Butcher series stems from their use to derive and to analyze numerical methods for differential equations. For example, consider the <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107076.png" />-stage Runge–Kutta formula (cf. [[Runge–Kutta method|Runge–Kutta method]])
+
To define  $  \psi ( t ) $,
 +
first let
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107077.png" /></td> <td valign="top" style="width:5%;text-align:right;">(a2)</td></tr></table>
+
$$
 +
\phi ( \emptyset ) = ( 1 \dots 1 ) ^ {T} \in \mathbf R  ^ {s} ,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107078.png" /></td> </tr></table>
+
$$
 +
\phi ( \bullet ) = A \phi ( \emptyset ) ,
 +
$$
  
for the ordinary differential equation <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107079.png" />. Let <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107080.png" /> be the vector of weights and <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107081.png" /> the coefficient matrix for (a2). Then it can be shown that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107082.png" /> can be written as the Butcher series
+
and then, for $  t = [ t _ {1} \dots t _ {m} ] $
 +
with  $  \rho ( t _ {i} ) \geq  1 $
 +
for  $  i = 1 \dots m $,
 +
define  $  \phi ( t ) $
 +
recursively by
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107083.png" /></td> <td valign="top" style="width:5%;text-align:right;">(a3)</td></tr></table>
+
$$
 +
\phi ( t ) = \rho ( t ) A ( \phi ( t _ {1} ) \dots \phi ( t _ {m} ) ) ,
 +
$$
  
where the coefficients <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107084.png" />, defined below, depend only on the rooted tree <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107085.png" /> and the coefficients of the formula (a2).
+
where the product of $  \phi $'s in the bracket is taken component-wise. Then
  
To define <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107086.png" />, first let
+
$$
 +
\psi ( \emptyset ) = 1,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107087.png" /></td> </tr></table>
+
$$
 +
\psi ( \bullet ) = \sum _ {i = 1 } ^ { s }  b _ {i} ,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107088.png" /></td> </tr></table>
+
and, for  $  t = [ t _ {1} \dots t _ {m} ] $
 +
with  $  \rho ( t _ {i} ) \geq  1 $
 +
for  $  i = 1 \dots m $,
  
and then, for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107089.png" /> with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107090.png" /> for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107091.png" />, define <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107092.png" /> recursively by
+
$$
 +
\psi ( t ) = \rho ( t ) b ^ {T} ( \phi ( t _ {1} ) \dots \phi ( t _ {m} ) ) ,
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107093.png" /></td> </tr></table>
+
where, again, the product of  $  \phi $'s in the bracket is taken componentwise.
  
where the product of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107094.png" />'s in the bracket is taken component-wise. Then
+
Assuming that  $  y ( x ) = y _ {n} $
 +
and that  $  \psi ( t ) = 1 $
 +
for all trees  $  t $
 +
of order  $  \leq  p $,
 +
it follows immediately from \eqref{eq:1} and \eqref{eq:3} that
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107095.png" /></td> </tr></table>
+
$$
 +
y _ {n + 1 }  = y ( x + h ) + {\mathcal O} ( h ^ {p + 1 } ) .
 +
$$
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107096.png" /></td> </tr></table>
+
The order of \eqref{eq:2} is the largest such $p$.
  
and, for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107097.png" /> with <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107098.png" /> for <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b11107099.png" />,
+
====Algebraic aspects====
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b111070100.png" /></td> </tr></table>
+
The Butcher series form a group, as they can be composed. This group is sometimes considered from the viewpoint of its Hopf algebra of functions.
  
where, again, the product of <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b111070101.png" />'s in the bracket is taken componentwise.
+
The modern understanding of this notion uses the notion of pre-Lie algebras, for the following reasons:
  
Assuming that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b111070102.png" /> and that <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b111070103.png" /> for all trees <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b111070104.png" /> of order <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b111070105.png" />, it follows immediately from (a1) and (a3) that
+
- The Lie algebra of analytic vector fields on the affine space $\RR^d$ comes by anti-symmetrization from the pre-Lie product given by the usual torsion-free and flat connection.
  
<table class="eq" style="width:100%;"> <tr><td valign="top" style="width:94%;text-align:center;"><img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b111070106.png" /></td> </tr></table>
+
- The free pre-Lie algebra on one generator can be described as the vector space spanned by rooted trees, under some kind of grafting.
  
The order of (a2) is the largest such <img align="absmiddle" border="0" src="https://www.encyclopediaofmath.org/legacyimages/b/b111/b111070/b111070107.png" />.
+
Then, by the universal property of free pre-Lie algebras, the choice of any analytic vector field defines uniquely a morphism of pre-Lie algebras, which gives the "elementary differentials" associating a vector field to each rooted tree.
  
 
====References====
 
====References====
<table><TR><TD valign="top">[a1]</TD> <TD valign="top">  J.C. Butcher,  "The numerical analysis of ordinary differential equations" , Wiley  (1987)</TD></TR><TR><TD valign="top">[a2]</TD> <TD valign="top">  E. Hairer,  S.P. Nørsett,  G. Wanner,  "Solving ordinary differential equations I: nonstiff problems" , Springer  (1987)</TD></TR></table>
+
<table>
 +
<TR><TD valign="top">[a1]</TD> <TD valign="top">  J.C. Butcher,  "The numerical analysis of ordinary differential equations" , Wiley  (1987) {{ZBL|0616.65072}}</TD></TR>
 +
<TR><TD valign="top">[a2]</TD> <TD valign="top">  E. Hairer,  S.P. Nørsett,  G. Wanner,  "Solving ordinary differential equations I: nonstiff problems" , Springer  (1987) {{ZBL|0638.65058}}</TD></TR>
 +
</table>

Latest revision as of 08:28, 12 November 2023


B-series

Let $ y : \mathbf R \rightarrow {\mathbf R ^ {d} } $ be an analytic function satisfying an ordinary differential equation $ y ^ \prime ( x ) = f ( y ( x ) ) $. A Butcher series for $ y $ is a series of the form

$$ B ( a,y ( x ) ) = \sum _ {t \in T } a ( t ) F ( t ) ( y ( x ) ) { \frac{h ^ {\rho ( t ) } }{\rho ( t ) ! } } , $$

where $ T $ is the set of rooted trees, $ a ( t ) \in \mathbf R $ is the coefficient for the series associated with the tree $ t $, $ F ( t ) ( y ( x ) ) \in \mathbf R ^ {d} $ is the elementary differential associated with the tree $ t $, $ h \in \mathbf R $ is a stepsize or $ x $- increment, and $ \rho ( t ) $, often called the order of the tree $ t $, is the number of nodes in the tree $ t $. All terms will be defined below. For a more thorough discussion of Butcher series, see [a1] or [a2].

Definitions

A tree is a connected acyclic graph (cf. also Graph). A rooted tree is a tree with a distinguished node, called the root (which is drawn at the bottom of the sample trees in the Table). Two rooted trees $t$ and $\widehat{t}$ are isomorphic (i.e. equivalent) if and only if there is a bijection $\sigma $ from the nodes of $t$ to the nodes of $\widehat{t}$ such that $ \sigma $ maps the root of $ t $ to the root of $ {\widehat{t} } $ and any two nodes $ n _ {1} $ and $ n _ {2} $ are connected by an edge in $ t $ if and only if $ \sigma ( n _ {1} ) $ and $ \sigma ( n _ {2} ) $ are connected by an edge in $ {\widehat{t} } $.

The following bracket notation for rooted trees is very useful. Let $\bullet$ be the tree with one node. For rooted trees $ t_{1} \dots t _ {m} $ with $ \rho ( t _ {i} ) \geq 1 $ for $ i = 1 \dots m $, let $ t = [ t _ {1} \dots t _ {m} ] $ be the rooted tree formed by connecting a new root to the roots of each of $ t _ {1} \dots t _ {m} $. Clearly, $ \rho ( t ) = 1 + \sum _ {i = 1 } ^ {m} \rho ( t _ {i} ) $ and any rooted tree formed by permuting the subtrees $ t _ {1} \dots t _ {m} $ of $ t $ is isomorphic to $ t $.

The rooted trees with at most four nodes are displayed in the table below.

$t_1$ $t_2=[t_1]$ $ t_{3} = [t_{1},t_{1}]$ $t_4 = [t_2]$
Rooted tree 0.svg Rooted tree 10.svg Rooted tree 200.svg Rooted tree 110.svg
$t_5 =[t_1,t_1,t_1]$ $t_6=[t_2,t_1]$ $ t_{7} = [t_{3}]$ $ t_{8} = [t_4]$
Rooted tree 3000.svg Rooted tree 2100.svg Rooted tree 1200.svg Rooted tree 1110.svg

Using this bracket notation for rooted trees, one can recursively define elementary differentials, as follows. Let

$$ F ( \emptyset ) ( y ( x ) ) = y ( x ) , \quad F ( \bullet ) ( y ( x ) ) = f ( y ( x ) ) , $$

and, for $ t = [ t _ {1} \dots t _ {m} ] $ with $ \rho ( t _ {i} ) \geq 1 $ for $ i = 1 \dots m $,

$$ F ( t ) ( y ( x ) ) = $$

$$ = f ^ {( m ) } ( y ( x ) ) ( F ( t _ {1} ) ( y ( x ) ) \dots F ( t _ {m} ) ( y ( x ) ) ) , $$

where $ f ^ {( m ) } ( y ( x ) ) $, the $ m $ th derivative of $ f $ with respect to $ y $, is a multi-linear mapping.

Since $ y ( x ) $ satisfies $ y ^ \prime ( x ) = f ( y ( x ) ) $, it is not hard to show that

$$ y ^ {( i ) } ( x ) = \sum _ {t \in T _ {i} } \alpha ( t ) F ( t ) ( y ( x ) ) , $$

where $ y ^ {( i ) } ( x ) $ is the $ i $ th derivative of $ y $ with respect to $ x $, $ T _ {i} = \{ {t \in T } : {\rho ( t ) = i } \} $, and $ \alpha ( t ) $ is the number of distinct ways of labeling the nodes of $ t $ with the integers $ \{ 1 \dots \rho ( t ) \} $ such that the labels increase as you follow any path from the root to a leaf of $ t $. Consequently, one can write the Taylor series for $ y ( x ) $ as a Butcher series:

\begin{equation} \label{eq:1} y ( x + h ) = \sum _ {i = 0 } ^ \infty y ^ {( i ) } ( x ) { \frac{h ^ {i} }{i! } } = \end{equation}

$$ = \sum _ {t \in T } \alpha ( t ) F ( t ) ( y ( x ) ) { \frac{h ^ {\rho ( t ) } }{\rho ( t ) ! } } . $$

The importance of Butcher series stems from their use to derive and to analyze numerical methods for differential equations. For example, consider the $ s $- stage Runge–Kutta formula (cf. Runge–Kutta method)

\begin{equation}\label{eq:2} y _ {n + 1 } = y _ {n} + h \sum _ {i = 1 } ^ { s } b _ {i} k _ {i} , \end{equation}

$$ k _ {i} = f \left ( y _ {n} + h \sum _ {j = 1 } ^ { s } a _ {ij } k _ {j} \right ) , $$

for the ordinary differential equation $ y ^ \prime ( x ) = f ( y ( x ) ) $. Let $ b = ( b _ {1} \dots b _ {s} ) ^ {T} \in \mathbf R ^ {s} $ be the vector of weights and $ A = [ a _ {ij } ] \in \mathbf R ^ {s \times s } $ the coefficient matrix for (a2). Then it can be shown that $ y _ {n + 1 } $ can be written as the Butcher series

\begin{equation}\label{eq:3} y _ {n + 1 } = \sum _ {t \in T } \alpha ( t ) \psi ( t ) F ( t ) ( y _ {n} ) { \frac{h ^ {\rho ( t ) } }{\rho ( t ) ! } } , \end{equation}

where the coefficients $ \psi(t) \in \mathbf R $, defined below, depend only on the rooted tree $t$ and the coefficients of the formula \eqref{eq:2}.

To define $ \psi ( t ) $, first let

$$ \phi ( \emptyset ) = ( 1 \dots 1 ) ^ {T} \in \mathbf R ^ {s} , $$

$$ \phi ( \bullet ) = A \phi ( \emptyset ) , $$

and then, for $ t = [ t _ {1} \dots t _ {m} ] $ with $ \rho ( t _ {i} ) \geq 1 $ for $ i = 1 \dots m $, define $ \phi ( t ) $ recursively by

$$ \phi ( t ) = \rho ( t ) A ( \phi ( t _ {1} ) \dots \phi ( t _ {m} ) ) , $$

where the product of $ \phi $'s in the bracket is taken component-wise. Then

$$ \psi ( \emptyset ) = 1, $$

$$ \psi ( \bullet ) = \sum _ {i = 1 } ^ { s } b _ {i} , $$

and, for $ t = [ t _ {1} \dots t _ {m} ] $ with $ \rho ( t _ {i} ) \geq 1 $ for $ i = 1 \dots m $,

$$ \psi ( t ) = \rho ( t ) b ^ {T} ( \phi ( t _ {1} ) \dots \phi ( t _ {m} ) ) , $$

where, again, the product of $ \phi $'s in the bracket is taken componentwise.

Assuming that $ y ( x ) = y _ {n} $ and that $ \psi ( t ) = 1 $ for all trees $ t $ of order $ \leq p $, it follows immediately from \eqref{eq:1} and \eqref{eq:3} that

$$ y _ {n + 1 } = y ( x + h ) + {\mathcal O} ( h ^ {p + 1 } ) . $$

The order of \eqref{eq:2} is the largest such $p$.

Algebraic aspects

The Butcher series form a group, as they can be composed. This group is sometimes considered from the viewpoint of its Hopf algebra of functions.

The modern understanding of this notion uses the notion of pre-Lie algebras, for the following reasons:

- The Lie algebra of analytic vector fields on the affine space $\RR^d$ comes by anti-symmetrization from the pre-Lie product given by the usual torsion-free and flat connection.

- The free pre-Lie algebra on one generator can be described as the vector space spanned by rooted trees, under some kind of grafting.

Then, by the universal property of free pre-Lie algebras, the choice of any analytic vector field defines uniquely a morphism of pre-Lie algebras, which gives the "elementary differentials" associating a vector field to each rooted tree.

References

[a1] J.C. Butcher, "The numerical analysis of ordinary differential equations" , Wiley (1987) Zbl 0616.65072
[a2] E. Hairer, S.P. Nørsett, G. Wanner, "Solving ordinary differential equations I: nonstiff problems" , Springer (1987) Zbl 0638.65058
How to Cite This Entry:
Butcher series. Encyclopedia of Mathematics. URL: http://encyclopediaofmath.org/index.php?title=Butcher_series&oldid=16379
This article was adapted from an original article by K.R. Jackson (originator), which appeared in Encyclopedia of Mathematics - ISBN 1402006098. See original article