Vector fields in vector spaces #
We study functions of the form V : E โ E
on a vector space, thinking of these as vector fields.
We define several notions in this context, with the aim to generalize them to vector fields on
manifolds.
Notably, we define the pullback of a vector field under a map, as
VectorField.pullback ๐ f V x := (fderiv ๐ f x).inverse (V (f x))
(together with the same notion
within a set).
We also define the Lie bracket of two vector fields as
VectorField.lieBracket ๐ V W x := fderiv ๐ W x (V x) - fderiv ๐ V x (W x)
(together with the same notion within a set).
In addition to comprehensive API on these two notions, the main results are the following:
VectorField.pullback_lieBracket
states that the pullback of the Lie bracket is the Lie bracket of the pullbacks, when the second derivative is symmetric.VectorField.leibniz_identity_lieBracket
is the Leibniz identity[U, [V, W]] = [[U, V], W] + [V, [U, W]]
.
The Lie bracket of vector fields in a vector space #
We define the Lie bracket of two vector fields, and call it lieBracket ๐ V W x
. We also define
a version localized to sets, lieBracketWithin ๐ V W s x
. We copy the relevant API
of fderivWithin
and fderiv
for these notions to get a comprehensive API.
The Lie bracket [V, W] (x)
of two vector fields at a point, defined as
DW(x) (V x) - DV(x) (W x)
.
Equations
- VectorField.lieBracket ๐ V W x = (fderiv ๐ W x) (V x) - (fderiv ๐ V x) (W x)
Instances For
The Lie bracket [V, W] (x)
of two vector fields within a set at a point, defined as
DW(x) (V x) - DV(x) (W x)
where the derivatives are taken inside s
.
Equations
- VectorField.lieBracketWithin ๐ V W s x = (fderivWithin ๐ W s x) (V x) - (fderivWithin ๐ V s x) (W x)
Instances For
Variant of lieBracketWithin_congr_set
where one requires the sets to coincide only in
the complement of a point.
Variant of lieBracketWithin_eventually_congr_set
where one requires the sets to coincide only
in the complement of a point.
If vector fields coincide on a neighborhood of a point within a set, then the Lie brackets also coincide on a neighborhood of this point within this set. Version where one considers the Lie bracket within a subset.
Version of lieBracketWithin_congr
in which one assumes that the point belongs to the
given set.
The Lie bracket of vector fields in vector spaces satisfies the Leibniz identity
[U, [V, W]] = [[U, V], W] + [V, [U, W]]
.
The Lie bracket of vector fields in vector spaces satisfies the Leibniz identity
[U, [V, W]] = [[U, V], W] + [V, [U, W]]
.
The Lie bracket of vector fields in vector spaces satisfies the Leibniz identity
[U, [V, W]] = [[U, V], W] + [V, [U, W]]
.
The pullback of vector fields in a vector space #
The pullback of a vector field under a function, defined
as (f^* V) (x) = Df(x)^{-1} (V (f x))
. If Df(x)
is not invertible, we use the junk value 0
.
Equations
- VectorField.pullback ๐ f V x = (fderiv ๐ f x).inverse (V (f x))
Instances For
The pullback within a set of a vector field under a function, defined
as (f^* V) (x) = Df(x)^{-1} (V (f x))
where Df(x)
is the derivative of f
within s
.
If Df(x)
is not invertible, we use the junk value 0
.
Equations
- VectorField.pullbackWithin ๐ f V s x = (fderivWithin ๐ f s x).inverse (V (f x))
Instances For
If a C^2
map has an invertible derivative within a set at a point, then nearby derivatives
can be written as continuous linear equivs, which depend in a C^1
way on the point, as well as
their inverse, and moreover one can compute the derivative of the inverse.
If a C^2
map has an invertible derivative at a point, then nearby derivatives can be written
as continuous linear equivs, which depend in a C^1
way on the point, as well as their inverse, and
moreover one can compute the derivative of the inverse.
The Lie bracket commutes with taking pullbacks. This requires the function to have symmetric
second derivative. Version in a complete space. One could also give a version avoiding
completeness but requiring that f
is a local diffeo.
The Lie bracket commutes with taking pullbacks. This requires the function to have symmetric
second derivative. Version in a complete space. One could also give a version avoiding
completeness but requiring that f
is a local diffeo. Variant where unique differentiability and
the invariance property are only required in a smaller set u
.
The Lie bracket commutes with taking pullbacks. This requires the function to have symmetric
second derivative. Version in a complete space. One could also give a version avoiding
completeness but requiring that f
is a local diffeo.
The Lie bracket commutes with taking pullbacks. This requires the function to have symmetric
second derivative. Version in a complete space. One could also give a version avoiding
completeness but requiring that f
is a local diffeo.
The Lie bracket commutes with taking pullbacks. One could also give a version avoiding
completeness but requiring that f
is a local diffeo.