completely positive map
{{Short description|C*-algebra mapping preserving positive elements}}
{{Use American English|date=January 2019}}
{{More citations needed|date=July 2020}}
In mathematics a positive map is a map between C*-algebras that sends positive elements to positive elements. A completely positive map is one that satisfies a stronger, more robust condition.
Definition
Let and be C*-algebras. A linear map is called a positive map if maps positive elements to positive elements: .
Any linear map induces another map
:
in a natural way. If is identified with the C*-algebra of -matrices with entries in , then acts as
:
\begin{pmatrix}
a_{11} & \cdots & a_{1k} \\
\vdots & \ddots & \vdots \\
a_{k1} & \cdots & a_{kk}
\end{pmatrix} \mapsto \begin{pmatrix}
\phi(a_{11}) & \cdots & \phi(a_{1k}) \\
\vdots & \ddots & \vdots \\
\phi(a_{k1}) & \cdots & \phi(a_{kk})
\end{pmatrix}.
We then say is k-positive if is a positive map and completely positive if is k-positive for all k.
Properties
- Positive maps are monotone, i.e. for all self-adjoint elements .
- Since for all self-adjoint elements , every positive map is automatically continuous with respect to the C*-norms and its operator norm equals . A similar statement with approximate units holds for non-unital algebras.
- The set of positive functionals is the dual cone of the cone of positive elements of .
Examples
- Every *-homomorphism is completely positive.K. R. Davidson: C*-Algebras by Example, American Mathematical Society (1996), ISBN 0-821-80599-1, Thm. IX.4.1
- For every linear operator between Hilbert spaces, the map is completely positive.R.V. Kadison, J. R. Ringrose: Fundamentals of the Theory of Operator Algebras II, Academic Press (1983), ISBN 0-1239-3302-1, Sect. 11.5.21 Stinespring's theorem says that all completely positive maps are compositions of *-homomorphisms and these special maps.
- Every positive functional (in particular every state) is automatically completely positive.
- Given the algebras and of complex-valued continuous functions on compact Hausdorff spaces , every positive map is completely positive.
- The transposition of matrices is a standard example of a positive map that fails to be 2-positive. Let {{math|T}} denote this map on . The following is a positive matrix in :
\begin{bmatrix}
\begin{pmatrix}1&0\\0&0\end{pmatrix}&
\begin{pmatrix}0&1\\0&0\end{pmatrix}\\
\begin{pmatrix}0&0\\1&0\end{pmatrix}&
\begin{pmatrix}0&0\\0&1\end{pmatrix}
\end{bmatrix}
=
\begin{bmatrix}
1 & 0 & 0 & 1 \\
0 & 0 & 0 & 0 \\
0 & 0 & 0 & 0 \\
1 & 0 & 0 & 1 \\
\end{bmatrix}.
The image of this matrix under is
\begin{bmatrix}
\begin{pmatrix}1&0\\0&0\end{pmatrix}^T&
\begin{pmatrix}0&1\\0&0\end{pmatrix}^T\\
\begin{pmatrix}0&0\\1&0\end{pmatrix}^T&
\begin{pmatrix}0&0\\0&1\end{pmatrix}^T
\end{bmatrix}
=
\begin{bmatrix}
1 & 0 & 0 & 0 \\
0 & 0 & 1 & 0 \\
0 & 1 & 0 & 0 \\
0 & 0 & 0 & 1 \\
\end{bmatrix} ,
which is clearly not positive, having determinant −1. Moreover, the eigenvalues of this matrix are 1,1,1 and −1. (This matrix happens to be the Choi matrix of T, in fact.) {{pb}} Incidentally, a map Φ is said to be co-positive if the composition Φ T is positive. The transposition map itself is a co-positive map.
See also
References
{{reflist}}