This week I have been reviewing the new spinfoam vertex in 4d models of quantum gravity. This was discussed in the recent posts:

- Quantum geometry from phase space reduction by Conrady and Freidel
- Holomorphic Factorization for a Quantum Tetrahedron by Freidel, Krasnov and Livine
- A new spinfoam vertex for quantum gravity by Livine and Speziale

In this post I explore the large spins asymptotic properties of the overlap coefficients:

characterizing the holomorphic intertwiners in the usual real basis. This consists of the normalization coefficient times the shifted Jacobi polynomial.

In the case of n = 4. I can study the asymptotics of the shifted Jacobi polynomials in the limit ji → λji, λ → ∞. A convenient integral representation for the shifted Jacobi polynomials is given by a contour integral:

This leads to the result that:

This formula relates the two very different descriptions of the phase space of shapes of a classical tetrahedron – the real one in terms of the k, φ parameters and the complex one in terms of the cross-ratio

coordinate Z. As is clear from this formula, the relation between the two descriptions is non-trivial.

In this post I have only worked with the simplest case of this relation when all areas are equal. In this ‘equi-area‘ case where all four representations are equal ji = j, ∀i = 1, 2, 3, 4, as described in the post: Holomorphic Factorization for a Quantum Tetrahedron the overlap function is;

###### where

###### Θ=cosh−¹(√z)

###### kc=2jtanh(Re(Θ))

###### and

###### φc = π/2.

###### Sagemath program

Using sagemath I am able to evaluate the overlap coefficients for various values of j and the cross-ratios z.

###### Numerical results

Here I plot the modulus of the equi-area case state Ck, for j = 20, as a function of the spin label k, for the value of the cross-ratio Z = exp(iπ/3) that corresponds to the equilateral tetrahedron. It is obvious that the distribution looks Gaussian. We also see that the maximum is reached for kc = 2j/√3 ∼ 23, which agrees with an asymptotic analysis.

Here I plot the modulus of the equi-area case state Ck for various j values as a function of the spin label k, for the value of the cross-ratio Z = exp(iπ/3) that corresponds to the equilateral tetrahedron.

Here I have have plotted the modulus of the j = 20 equi-area state Ck for increasing cross-ratios Z = 0.1i, 0.8i, 1.8i. The Gaussian distribution progressively moving its peak from 0 to 2j. This illustrates how changing the value of Z affects the semi-classical geometry of the tetrahedron.

**Conclusions**

In this post I we have studied a holomorphic basis for the Hilbert space Hj1,…,jn of SU(2) intertwiners. In particular I have looked at the case of 4-valent intertwiners that can be interpreted as quantum states of a quantum tetrahedron. The formula

gives the inner product in Hj1,…,jn in terms of a holomorphic integral over the space of ‘shapes’ parametrized by the cross-ratio coordinates Zi. In the tetrahedral n = 4 case there is a single cross-ratio Z. The n=4 holomorphic intertwiners parametrized by a single cross-ratio variable Z are coherent states in that they form an over-complete basis of the Hilbert space of intertwiners and are semi-classical states peaked on the geometry of a classical tetrahedron as shown by my numerical studies. The new holomorphic intertwiners are related to the standard spin basis of intertwiners that are usually used in loop quantum gravity and spin foam models, and the change of basis coefficients are given by Jacobi polynomials.

In the canonical framework of loop quantum gravity, spin network states of quantum geometry are labeled by a graph as well as by SU(2) representations on the graph’s edges e and intertwiners on its vertices v. It is now possible to put holomorphic intertwiners at the vertices of the graph, which introduces the new spin networks labeled by representations je and cross-ratios Zv. Since each holomorphic intertwiner can be associated to a classical tetrahedron, we can interpret these new spin network states as discrete geometries. In particular, geometrical observables such as the volume can be expected to be peaked on their classical values as shown in my numerical studies for j=20. This should be of great help when looking at the dynamics of the spin network states and when studying how they are coarse-grained and refined.

###### Related articles

- The SageMath Cloud – computational mathematics in the cloud (jugad2.blogspot.com)
- A new spinfoam vertex for quantum gravity by Livine and Speziale (quantumtetrahedron.wordpress.com)
- Quantum geometry from phase space reduction by Conrady and Freidel (quantumtetrahedron.wordpress.com)
- Holomorphic Factorization for a Quantum Tetrahedron by Freidel, Krasnov and Livine (quantumtetrahedron.wordpress.com)
- Advanced Analysis, Notes 19: The holomorphic functional calculus II (definition and basic properties) (noncommutativeanalysis.wordpress.com)
- K-spectral sets and the holomorphic functional calculus (noncommutativeanalysis.wordpress.com)

## One thought on “Numerical work with sagemath 15: Holomorphic factorization”