in quantum computation

Similar documents
it is easy to see that α = a

Finite dimensional C -algebras

NOTES ON CATEGORIES AND FUNCTORS

Chapter 13: Basic ring theory

SOME PROPERTIES OF FIBER PRODUCT PRESERVING BUNDLE FUNCTORS

Section Inner Products and Norms

Mathematics Course 111: Algebra I Part IV: Vector Spaces

ON GALOIS REALIZATIONS OF THE 2-COVERABLE SYMMETRIC AND ALTERNATING GROUPS

3. Prime and maximal ideals Definitions and Examples.

FIBER PRODUCTS AND ZARISKI SHEAVES

How To Prove The Dirichlet Unit Theorem

(a) Write each of p and q as a polynomial in x with coefficients in Z[y, z]. deg(p) = 7 deg(q) = 9

GROUPS ACTING ON A SET

1 = (a 0 + b 0 α) (a m 1 + b m 1 α) 2. for certain elements a 0,..., a m 1, b 0,..., b m 1 of F. Multiplying out, we obtain

Finite Projective demorgan Algebras. honoring Jorge Martínez

1 Sets and Set Notation.

The Division Algorithm for Polynomials Handout Monday March 5, 2012

ISOMETRIES OF R n KEITH CONRAD

Online publication date: 11 March 2010 PLEASE SCROLL DOWN FOR ARTICLE

calculating the result modulo 3, as follows: p(0) = = 1 0,

Parametric Domain-theoretic models of Linear Abadi & Plotkin Logic

Let H and J be as in the above lemma. The result of the lemma shows that the integral

Polynomial Invariants

EXERCISES FOR THE COURSE MATH 570, FALL 2010

Finite dimensional topological vector spaces

ACTA UNIVERSITATIS APULENSIS No 15/2008 PRODUCTS OF MULTIALGEBRAS AND THEIR FUNDAMENTAL ALGEBRAS. Cosmin Pelea

Lecture 18 - Clifford Algebras and Spin groups

Introduction to Modern Algebra

= = 3 4, Now assume that P (k) is true for some fixed k 2. This means that

ALGEBRA HW 5 CLAY SHONKWILER

FUNCTIONAL ANALYSIS LECTURE NOTES: QUOTIENT SPACES

Math 312 Homework 1 Solutions

On the structure of C -algebra generated by a family of partial isometries and multipliers

Chapter 4, Arithmetic in F [x] Polynomial arithmetic and the division algorithm.

RESULTANT AND DISCRIMINANT OF POLYNOMIALS

INTRODUCTORY SET THEORY

Linear Maps. Isaiah Lankham, Bruno Nachtergaele, Anne Schilling (February 5, 2007)

Similarity and Diagonalization. Similar Matrices

BANACH AND HILBERT SPACE REVIEW

Domino Tableaux, Schützenberger Involution, and the Symmetric Group Action

Group Theory. Contents

PROBLEM SET 6: POLYNOMIALS

Quotient Rings and Field Extensions

Factorization in Polynomial Rings

LINEAR ALGEBRA W W L CHEN

Non-unique factorization of polynomials over residue class rings of the integers

GENERATING SETS KEITH CONRAD

T ( a i x i ) = a i T (x i ).

POLYNOMIAL RINGS AND UNIQUE FACTORIZATION DOMAINS

MATH PROBLEMS, WITH SOLUTIONS

I. GROUPS: BASIC DEFINITIONS AND EXAMPLES

4: EIGENVALUES, EIGENVECTORS, DIAGONALIZATION

THE FUNDAMENTAL THEOREM OF ALGEBRA VIA PROPER MAPS

4.1 Modules, Homomorphisms, and Exact Sequences

CODING TRUE ARITHMETIC IN THE MEDVEDEV AND MUCHNIK DEGREES

The Convolution Operation

Chapter 7. Homotopy. 7.1 Basic concepts of homotopy. Example: z dz. z dz = but

GROUP ALGEBRAS. ANDREI YAFAEV

Linear Types for Continuations

Linear Algebra. A vector space (over R) is an ordered quadruple. such that V is a set; 0 V ; and the following eight axioms hold:

Direct models of the computational lambda-calculus

Mathematics for Computer Science/Software Engineering. Notes for the course MSM1F3 Dr. R. A. Wilson

Foundational Methods in Computer Science 2013 Abstracts

Factorization Algorithms for Polynomials over Finite Fields

BABY VERMA MODULES FOR RATIONAL CHEREDNIK ALGEBRAS

Techniques algébriques en calcul quantique

JET-LIKE APPROACH TO WEIL BUNDLES

RINGS WITH A POLYNOMIAL IDENTITY

1 Homework 1. [p 0 q i+j p i 1 q j+1 ] + [p i q j ] + [p i+1 q j p i+j q 0 ]

SMALL SKEW FIELDS CÉDRIC MILLIET

MATH 304 Linear Algebra Lecture 9: Subspaces of vector spaces (continued). Span. Spanning set.

ON SOME CLASSES OF REGULAR ORDER SEMIGROUPS

MATH 304 Linear Algebra Lecture 20: Inner product spaces. Orthogonal sets.

9. Quotient Groups Given a group G and a subgroup H, under what circumstances can we find a homomorphism φ: G G ', such that H is the kernel of φ?

Point Set Topology. A. Topological Spaces and Continuous Maps

Row Ideals and Fibers of Morphisms

Chapter 7. Permutation Groups

How To Find Out How To Build An Elliptic Curve Over A Number Field

Matrix Representations of Linear Transformations and Changes of Coordinates

MATRIX ALGEBRA AND SYSTEMS OF EQUATIONS

IRREDUCIBLE OPERATOR SEMIGROUPS SUCH THAT AB AND BA ARE PROPORTIONAL. 1. Introduction

ABSTRACT ALGEBRA: A STUDY GUIDE FOR BEGINNERS

MATH 101A: ALGEBRA I PART B: RINGS AND MODULES

16.3 Fredholm Operators

Introduction to Finite Fields (cont.)

Inner Product Spaces and Orthogonality

A NOTE ON TRIVIAL FIBRATIONS

Geometric Transformations

Sets of Fibre Homotopy Classes and Twisted Order Parameter Spaces

Module MA3411: Abstract Algebra Galois Theory Appendix Michaelmas Term 2013

1 VECTOR SPACES AND SUBSPACES

Computing the Symmetry Groups of the Platonic Solids With the Help of Maple

Generic Polynomials of Degree Three

NOTES ON LINEAR TRANSFORMATIONS

Copy in your notebook: Add an example of each term with the symbols used in algebra 2 if there are any.

CLUSTER ALGEBRAS AND CATEGORIFICATION TALKS: QUIVERS AND AUSLANDER-REITEN THEORY

Inner products on R n, and more

3 1. Note that all cubes solve it; therefore, there are no more

A numerable cover of a topological space X is one which possesses a partition of unity.

3. Let A and B be two n n orthogonal matrices. Then prove that AB and BA are both orthogonal matrices. Prove a similar result for unitary matrices.

Transcription:

in computation Kestrel Institute and Oxford University - Oxford, August 2008

Outline Quantum programming λ- Abstraction with pictures Consequences - Abstraction in Category of measurements -

Outline Quantum programming λ- Abstraction with pictures Consequences - Abstraction in Category of measurements Quantum programming λ- -

What do programmers do? Quantum programming λ- f x Z m 2 f(x) Z n 2 -

What do programmers do? x f x Z m 2 f(x) Z n 2 x Quantum programming λ- - y f(x) y

What do programmers do? Quantum programming λ- x x B m = C Zm 2 U f f - y f(x) y B n = C Zn 2

What do programmers do? Quantum programming λ- z H m x z = P x ( 1)z x x f x H m z x = P z ( 1)x z z - U f y f(x) y

What do programmers do? Simon s algorithm f : Z m 2 Zn 2 : x f(x) f : Z m+n 2 Z m+n 2 : x, y x, f(x) y U f : C Zm+n 2 C Zm+n 2 : x, y x, f(x) y Simon = (H m id)u f (H m id) 0, 0 = ( 1) x z z, f(x) Quantum programming λ- - x,z Z m 2

What do programmers do? Simon s algorithm f : Z m 2 Zn 2 : x f(x) f : Z m+n 2 Z m+n 2 : x, y x, f(x) y U f : C Zm+n 2 C Zm+n 2 : x, y x, f(x) y Simon = (H m id)u f (H m id) 0, 0 = ( 1) x z z, f(x) Quantum programming λ- - x,z Z m 2... to find a hidden subgroup measurement find c such that f(x + c) = f(x)

What do programmers do? Shor s algorithm f : Z m+n q U f : C Zm+n q f : Z m q Z n q : x a x mod q Z m+n q : x, y x, a x + y mod q C Zm+n q : x, y x, a x + y mod q Quantum programming λ- - Shor = (FT m id)u f (FT m id) 0, 0 = ( 1) x z z, f(x) x,z Z m q... to find a hidden subgroup measurement find c such that a x+c = a x mod q

What do programmers do? Hallgren s algorithm h : Z m Z n : x I x (fraction ideal) h : Z m+n Z m+n : x, y x, y h(x) U h : C Zm+n C Zm+n : x, y x, y h(x) Hallgren = (FT m id)u h (FT m id) d, d = x,z Z m ( 1) x z z, h(x) Quantum programming λ- -... to find a hidden subgroup measurement find R such that h(x + R) = h(x)

General design pattern of software engineering ;) CLASS QUAN T Quantum programming λ- - MEAS T CLASS

Quantum prog. = functional prog. + superposition + entanglement CLASS QUAN T superposition entanglement MEAS T Quantum programming λ- - CLASS

Standard type universes where programmers work CLASS FSet, FSet op, FFMod R... Quantum programming λ- QUAN T FHilb, CPM(FHilb)... - MEAS T FHilb, CPM(FHilb)... CLASS

Function in programming CLASS[x] QUAN T [x] FSet[x],FSet op [x], FFMod R[x]... FHilb[x], CPM(FHilb)[x]... Quantum programming λ- - MEAS T {x} FHilb{x}, CPM(FHilb){x}... CLASS

Function in programming CLASS[x] QUAN T [x] = KL (X ) = KL (X ) Quantum programming λ- - MEAS T {x} EM(X ) CLASS

λ- 2x 3 + 3x + 1 in Z[x] λx. 2x 3 + 3x + 1 in Z Z Z[x] x ad x Quantum programming λ- - Z f a f S a

λ- in cartesian closed categories S p(x) : B in S[x : X] λx. p(x) : B X in S S[x] 1 x X ad x F a Quantum programming λ- - F C 1 a FX

λ- in cartesian closed categories S A q(x) B in S[x : X] A λx.q(x) B X in S S[x] 1 x X ad x F a Quantum programming λ- - F C 1 a FX

λ- in cartesian closed categories A ϕ,x B X X ǫ B A ϕ B X S[x](A, B) S(A, B X ) A q(x) B A λx.q(x) B X S[x] 1 x X ad x Quantum programming λ- - S F a F C 1 a FX

λ- in cartesian closed categories Theorem (Lambek, Adv. in Math. 79) Let S be a cartesian closed category, and S[x] the free cartesian closed category generated by S and x : 1 X. Quantum programming λ- -

λ- in cartesian closed categories Theorem (Lambek, Adv. in Math. 79) Let S be a cartesian closed category, and S[x] the free cartesian closed category generated by S and x : 1 X. Then the inclusion ad x : S S[x] has a right adjoint ab x : S[x] S : A A X and the transpositions A ϕ,x B X X ǫ B A ϕ B X S[x](ad x A, B) S(A, ab x B) A q(x) B A λx.q(x) B X Quantum programming λ- - model λ- and application.

λ- in cartesian closed categories Theorem (Lambek, Adv. in Math. 79) Let S be a cartesian closed category, and S[x] the free cartesian closed category generated by S and x : 1 X. Then the inclusion ad x : S S[x] has a right adjoint ab x : S[x] S : A A X and the transpositions A ϕ,x B X X ǫ B A ϕ B X S[x](ad x A, B) S(A, ab x B) A q(x) B A λx.q(x) B X Quantum programming λ- - model λ- and application. S[x] is isomorphic with the Kleisli category for the power monad ( ) X.

κ- in cartesian Theorem (Lambek, Adv. in Math. 79) categories Let S be a cartesian category, and S[x] the free cartesian category generated by S and x : 1 X. Then the inclusion ad x : S S[x] has a left adjoint ab x : S[x] S : A X A and the transpositions A x,id X A ϕ B X A ϕ B S[x](A, ad x B) S(ab x A, B) A f x B X A κ x.fx B Quantum programming λ- - model first order and application. S[x] is isomorphic with the Kleisli category for the product comonad X ( ).

κ- in monoidal Theorem (DP, MSCS 95) categories Let C be a monoidal category, and C[x] the free monoidal category generated by C and x : 1 X. Then the strong adjunctions ab x ad x : C C[x] are in one-to-one correspondence with the internal comonoid structures on X. The transpositions A X A x A B ϕ C[x](A, ad x B) A f x B Quantum programming λ- - X A ϕ B C(ab x A, B) X A κ x.fx B model action and application. C[x] is isomorphic with the Kleisli category for the comonad X ( ), induced by any of the comonoid structures.

κ- in monoidal categories Task Extend this to. Quantum programming λ- -

κ- in monoidal categories Task Extend this to. Problem Lots of complicated diagram chasing. Quantum programming λ- -

κ- in monoidal categories Task Extend this to. Problem Lots of complicated diagram chasing. Solution? What does mean graphically? Quantum programming λ- -

Outline Quantum programming λ- Abstraction with pictures Consequences - Abstraction in Category of measurements -

Objects X A B D X A B D -

Identities X A B D X A B D X A B D X A B D id -

Morphisms B X B X h h - X A B D X A B D

Tensor (parallel composition) B X C B X C h f h f - X A B D X X A B D X

Sequential composition B X C B X C h f h f - X A B D X X A B D X g X A B g D D X X A B D D X

Elements (vectors) and coelements (functionals) b B X C h f B X B X b B X C h f - X A B D X X A B D X g X A B g a D D X X A B D D X X a D D X X I D D X

Symmetry b B X C h f B X c D g B X B X b B X C h f X A D B X X A c X X A B D X X A B g - X A a B D D X X A B D D X X a D D X X I D D X

Polynomials X b B X C h f B X c D g A B D D X a B X B X b B X C h f X A D B X id x X A D B I X A c r X A B D X A B g X A B D D X x a D D x I I D D I -

Outline Quantum programming λ- Abstraction with pictures Consequences - Abstraction in Category of measurements Abstraction with pictures Consequences -

Abstraction with pictures Theorem (again) Let C be a symmetric monoidal category, and C[x] the free symmetric monoidal category generated by C and x : 1 X. Then there is a one-to-one correspondence between adjunctions ab x ad x : C C[x] satisfying 1. ab x (A B) = ab x (A) B 2. η(a B) = η(a) B 3. η I = x and commutative comonoids on X. Abstraction with pictures Consequences -

Abstraction with pictures Theorem (again) Let C be a symmetric monoidal category, and C[x] the free symmetric monoidal category generated by C and x : 1 X. Then there is a one-to-one correspondence between adjunctions ab x ad x : C C[x] satisfying 1. ab x (A B) = ab x (A) B 2. η(a B) = η(a) B 3. η I = x and commutative comonoids on X. Abstraction with pictures Consequences - C[x] is isomorphic with the Kleisli category for the commutative comonad X ( ), induced by any of the comonoid structures.

Proof ( ) Given ab x ad x : C C[x], conditions 1.-3. imply ab x (A) = X A η(a) = x A Abstraction with pictures Consequences -

Proof ( ) Therefore the correspondence C ( ab x (A), B ) C[x] ( A, ad x (B) ) Abstraction with pictures Consequences -

Proof ( )... is actually C(X A, B) C[x](A, B) Abstraction with pictures Consequences -

Proof ( )... with C(X A, B) C[x](A, B) B f X A ( ) (x A) B f A Abstraction with pictures Consequences -

Proof ( )... and C(X A, B) C[x](A, B) h f κx. h f Abstraction with pictures Consequences - g g X

Proof ( ) The bijection corresponds to the conversion: ( ) (x A) C(X A, B) = C[x](A, B) κx. ( ) ( κx. ϕ(x) x A) = ϕ(x) (β-rule Abstraction with pictures Consequences - κx. ( f (x A) ) = f (η-rule

Proof ( ) The comonoid structure (X,, ) is = κx. = κx. X X id I Abstraction with pictures Consequences -

Proof ( ) The conversion rules imply the comonoid laws = = Abstraction with pictures Consequences - = =

Proof ( ) Given (X,, ), use its copying and deleting power, and the symmetries, to normalize every C[x]-arrow: ϕ(x) = ϕ (x A) h f h f Abstraction with pictures Consequences - g g

Proof ( ) Then set κx. ϕ(x) = ϕ to get ( ) (x A) C(X A, B) = C[x](A, B) κx. ( ) ( κx. ϕ(x) x A) = ϕ(x) (β-rule Abstraction with pictures Consequences - κx. ( f (x A) ) = f (η-rule

Remark C[x] = C X and C[x, y] = C X Y = KL(X Y ), reduce the finite polynomials to the Kleisli morphisms. Abstraction with pictures Consequences -

Remark C[x] = C X and C[x, y] = C X Y = KL(X Y ), reduce the finite polynomials to the Kleisli morphisms. But the extensions C[X], where X is large are also of interest. Abstraction with pictures Consequences -

Remark C[x] = C X and C[x, y] = C X Y = KL(X Y ), reduce the finite polynomials to the Kleisli morphisms. But the extensions C[X], where X is large are also of interest. Abstraction with pictures Consequences - Cf. N[N], Set[Set], and CPM(C).

Interpretation Abstraction with pictures Consequences -

Interpretation Upshot In symmetric monoidal categories, applies just to copiable and deletable data. Abstraction with pictures Consequences -

Interpretation Upshot In symmetric monoidal categories, applies just to copiable and deletable data. Definition A vector ϕ C(I, X) is a base vector (or a set-like element) with respect to the operation κx if it can be copied and deleted in C[x] (κx.x x) ϕ = ϕ ϕ Abstraction with pictures Consequences - (κx.id I ) ϕ = id I

Interpretation Upshot In symmetric monoidal categories, applies just to copiable and deletable data. Definition A vector ϕ C(I, X) is a base vector (or a set-like element) with respect to the operation κx if it can be copied and deleted in C[x] (κx.x x) ϕ = ϕ ϕ Abstraction with pictures Consequences - (κx.id I ) ϕ = id I Proposition ϕ C(I, X) is a base vector with respect to κx if and only if it is a homomorphism for the comonoid structure X X X I corresponding to κx.

Interpretation Upshot In other words, only the base vectors can be substituted for variables. Abstraction with pictures Consequences -

Interpretation Upshot In other words, only the base vectors can be substituted for variables. Definition Substitution is a structure preserving ioof C[x] C. Abstraction with pictures Consequences -

Interpretation Upshot In other words, only the base vectors can be substituted for variables. Definition Substitution is a structure preserving ioof C[x] C. Corollary The substitution functors C[x] C are in one-to-one correspondence with the base vectors of type X. Abstraction with pictures Consequences -

Outline Quantum programming λ- Abstraction with pictures Consequences - Abstraction in - Abstraction in Category of measurements

Definitions A -category C is given with an involutive ioof : C op C. - Abstraction in

Definitions A -category C is given with an involutive ioof : C op C. A morphism f in a -category C is called unitary if f = f 1. - Abstraction in

Definitions A -category C is given with an involutive ioof : C op C. A morphism f in a -category C is called unitary if f = f 1. A (symmetric) monoidal category C is -monoidal if its monoidal isomorphisms are unitary. - Abstraction in

Using the monoidal s for: vectors: C(A) = C(I, A) scalars: I = C(I, I) - Abstraction in

Using the monoidal s for: vectors: C(A) = C(I, A) scalars: I = C(I, I) in every -monoidal category we can define abstract inner product A : C(A) C(A) I ( ) ϕ (ϕ, ψ : I A) I A ψ I - Abstraction in

Using the monoidal s for: vectors: C(A) = C(I, A) scalars: I = C(I, I) in every -monoidal category we can define abstract inner product A : C(A) C(A) I ( ) ϕ (ϕ, ψ : I A) I A ψ I partial inner product AB : C(A B) C(A) C(B) ( ) (ϕ : I A B, ψ : I A) ϕ I A B ψ B B - Abstraction in

Using the monoidal s for: vectors: C(A) = C(I, A) scalars: I = C(I, I) in every -monoidal category we can define abstract inner product A : C(A) C(A) I ( ) ϕ (ϕ, ψ : I A) I A ψ I partial inner product AB : C(A B) C(A) C(B) ( ) (ϕ : I A B, ψ : I A) ϕ I A B ψ B B - Abstraction in entangled vectors η C(A A), such that ϕ C(A) η ϕ AA = ϕ

Using entangled vectors ηa : I A A and, ηb : I B Band their adjoints εa = η A : A A I and εb = η B : B B I - Abstraction in

Using entangled vectors ηa : I A A and, ηb : I B Band their adjoints εa = η A : A A I and εb = η B : B B I we can define for every f : A B the dual f : B A - Abstraction in f = B Bη BAA BfA BBA εa A

Using entangled vectors ηa : I A A and, ηb : I B Band their adjoints εa = η A : A A I and εb = η B : B B I we can define for every f : A B the dual f : B A - Abstraction in f = B Bη BAA BfA BBA εa A the conjugate f : A B f = f = f

Proposition For every object A in a -monoidal category C holds (a) (b) = (c), - Abstraction in

Proposition For every object A in a -monoidal category C holds (a) (b) = (c), where (a) η C(A A) is entangled - Abstraction in

Proposition For every object A in a -monoidal category C holds (a) (b) = (c), where (a) η C(A A) is entangled (b) ε = η C(A A, I) internalizes the inner product ε (ψ ϕ) = ϕ ψ - Abstraction in

Proposition For every object A in a -monoidal category C holds (a) (b) = (c), where (a) η C(A A) is entangled (b) ε = η C(A A, I) internalizes the inner product ε (ψ ϕ) = ϕ ψ (c) (η,ε) realize the self-adjunction A A, in the sense - Abstraction in A η A A A A A ε A = id A A A η A A A ε A A = id A The three conditions are equivalent if I generates C.

Proposition in pictures For every object A in a -monoidal category C holds (a) (b) = (c), where (a) (b) ψ η η ψ ϕ = = ψ ψ ϕ - Abstraction in (c) η η = = X η η

Definition A object in a -monoidal category is an object equipped with the structure from the preceding proposition. - Abstraction in

Definition A object in a -monoidal category is an object equipped with the structure from the preceding proposition. - Remark The subcategory of objects in any -monoidal category is -compact (strongly compact) with all objects self-adjoint. Abstraction in

Abstraction in Theorem Let C be a -monoidal category, and X X X I a comonoid that induces ab x ad x : C C[x]. - Abstraction in

Abstraction in Theorem Let C be a -monoidal category, and X X X I a comonoid that induces ab x ad x : C C[x]. Then the following conditions are equivalent: - Abstraction in

Abstraction in Theorem Let C be a -monoidal category, and X X X I a comonoid that induces ab x ad x : C C[x]. Then the following conditions are equivalent: (a) ad x : C C[x] creates : C[x] op C[x] such that x x = x x = id I. - Abstraction in

Abstraction in Theorem Let C be a -monoidal category, and X X X I a comonoid that induces ab x ad x : C C[x]. Then the following conditions are equivalent: (a) ad x : C C[x] creates : C[x] op C[x] such that x x = x x = id I. (b) η = and ε = η = realize X X. - Abstraction in

Abstraction in Theorem Let C be a -monoidal category, and X X X I a comonoid that induces ab x ad x : C C[x]. Then the following conditions are equivalent: (a) ad x : C C[x] creates : C[x] op C[x] such that x x = x x = id I. (b) η = and ε = η = realize X X. (c) (X ) ( X) = = ( X) (X ) - Abstraction in

Abstraction in Theorem Let C be a -monoidal category, and X X X I a comonoid that induces ab x ad x : C C[x]. Then the following conditions are equivalent: (a) ad x : C C[x] creates : C[x] op C[x] such that x x = x x = id I. (b) η = and ε = η = realize X X. (c) (X ) ( X) = = ( X) (X ) - Abstraction in where X X X I is the induced monoid = =

Abstraction in Theorem in pictures (b) = = X - Abstraction in

Abstraction in Theorem in pictures (c) = = - Abstraction in

Proof of (b)= (c) Lemma 1 If (b) holds then - = = Abstraction in

Proof of (b)= (c) Then (c) also holds because = = = - Abstraction in =

Proof of Lemma 1 Lemma 2 If η ε = = X ε η - Abstraction in then η ε ε = = X X ε ε η η η

Proof of Lemma 1 Using Lemma 2, and the fact that (b) implies = =, we get ε ε = η = η = - Abstraction in η η = η

The message of the proof - Abstraction in

The message of the proof There is more to categories than just diagram chasing. - Abstraction in

The message of the proof There is more to categories than just diagram chasing. - There is also picture chasing. Abstraction in

Definition A classical object in a -monoidal category C is a comonoid X X X I satisfying the equivalent conditions from the preceding theorem. - Abstraction in

Definition A classical object in a -monoidal category C is a comonoid X X X I satisfying the equivalent conditions from the preceding theorem. Let C be the category of classical objects and comonoid homomorphisms in C. - Abstraction in

Definition A classical object in a -monoidal category C is a comonoid X X X I satisfying the equivalent conditions from the preceding theorem. Let C be the category of classical objects and comonoid homomorphisms in C. Question: What is classical about classical objects? - Abstraction in

Definition A classical object in a -monoidal category C is a comonoid X X X I satisfying the equivalent conditions from the preceding theorem. Let C be the category of classical objects and comonoid homomorphisms in C. Question: What is classical about classical objects? classical structure structure - Abstraction in

Definition A classical object in a -monoidal category C is a comonoid X X X I satisfying the equivalent conditions from the preceding theorem. Let C be the category of classical objects and comonoid homomorphisms in C. Question: What is classical about classical objects? classical structure structure Answer: classical elements = base vectors - Abstraction in

Definition A classical object in a -monoidal category C is a comonoid X X X I satisfying the equivalent conditions from the preceding theorem. Let C be the category of classical objects and comonoid homomorphisms in C. Question: What is classical about classical objects? classical structure structure Answer: classical elements = base vectors is neither injective or surjective - Abstraction in

Consequences Upshot The Frobenius condition (c) assures the preservation of the operation under. - Abstraction in

Consequences Upshot The Frobenius condition (c) assures the preservation of the operation under. This leads to entanglement. - Abstraction in

Consequences Proposition The vectors C(X) of any classical object X form a -algebra. - Abstraction in

Consequences Proposition The vectors C(X) of any classical object X form a -algebra. - Abstraction in ϕ ψ = (ϕ ψ) ǫ = ϕ = ϕ = ϕ

Definition Two vectors ϕ, ψ C(A) in a -monoidal category are orthonormal if their inner product is idempotent: ϕ ψ = ϕ ψ 2 - Abstraction in

Definition Two vectors ϕ, ψ C(A) in a -monoidal category are orthonormal if their inner product is idempotent: ϕ ψ = ϕ ψ 2 - Proposition Any two base vectors are orthonormal. In particular, any two variables in a polynomial category are orthonormal. Abstraction in

Definition A classical object X is standard if it is (regularly) generated by its base vectors B(X) = {ϕ C(X) (κx. x x)ϕ = ϕ ϕ (κx. id I )ϕ = id I } - Abstraction in

Definition A classical object X is standard if it is (regularly) generated by its base vectors in the sense B(X) = {ϕ C(X) (κx. x x)ϕ = ϕ ϕ (κx. id I )ϕ = id I } f, g C(X, Y). ( ϕ B(X). fϕ = gϕ) = f = g - Abstraction in A base is regular if C(X, Y) C(Y) B(X) splits.

Proposition 1. All standard classical structures, that an object X C may carry, induce the bases with the same number of elements. - Abstraction in

Proposition 1. All standard classical structures, that an object X C may carry, induce the bases with the same number of elements. - Abstraction in Proposition 2. Let X C be a classical object with a regular base. Then the equipotent regular bases on any Y C are in one-to-one correspondence with the unitaries X Y.

Definition A qubit type in an arbitrary -monoidal category C is a classical object B with a unitary H of order 2. The induced bases are usually denoted by 0, 1, and +,. - Abstraction in

Definition A qubit type in an arbitrary -monoidal category C is a classical object B with a unitary H of order 2. The induced bases are usually denoted by 0, 1, and +,. - Computing with qubits A -monoidal category with B suffices for the basic algorithms. Abstraction in

Definition A qubit type in an arbitrary -monoidal category C is a classical object B with a unitary H of order 2. The induced bases are usually denoted by 0, 1, and +,. - Computing with qubits A -monoidal category with B suffices for the basic algorithms. A Klein group of unitaries on B suffices for all teleportation and dense coding schemes. Abstraction in

Proposition (Coecke, Vicary, P) Every classical object X in FHilb is regular, and X = C n. The classical structure is induced by a base B(X) = { i i n }, with i = ii = 1 n n i i=1 - Abstraction in

Proposition (Coecke, Vicary, P) Every classical object X in FHilb is regular, and X = C n. The classical structure is induced by a base B(X) = { i i n }, with i = ii = 1 n n i i=1 - Abstraction in Moreover, FHilb FSet

Proof A -algebra in FHilb is a C -algebra. - Abstraction in

Proof A -algebra in FHilb is a C -algebra. Thus for a classical X FHilb, : FHilb(X) FHilb(X, X) ( ) ϕ I X ( X ϕ X X X X ) is a representation of a commutative C -algebra. - Abstraction in

Proof Working through the Gelfand-Naimark duality, we get X = C n - Abstraction in

Proof Working through the Gelfand-Naimark duality, we get X = C n - Abstraction in because the spectrum of a commutative finitely dimensional C -algebra is a discrete set n of minimal central projections, while the representing spaces are the full matrix algebras C(1)

Outline Quantum programming λ- Abstraction with pictures Consequences - Abstraction in Category of measurements -

Category of measurements ((this was not presented)) -

Outline Quantum programming λ- Abstraction with pictures Consequences - Abstraction in Category of measurements -

Claim: Task: Simple algorithms have simple categorical semantics. Implement and analyze algorithms in nonstandard models: network computation, data mining. -