random_matrix

• full name: tenpy.linalg.random_matrix

• parent module: tenpy.linalg

• type: module

Functions

COE(size)	Circular orthogonal ensemble (COE).
CRE(size)	Circular real ensemble (CRE).
CUE(size)	Circular unitary ensemble (CUE).
GOE(size)	Gaussian orthogonal ensemble (GOE).
GUE(size)	Gaussian unitary ensemble (GUE).
O_close_1(size[, a])	return an random orthogonal matrix 'close' to the Identity.
U_close_1(size[, a])	return an random orthogonal matrix 'close' to the identity.
box(size[, W])	return random number uniform in (-W, W].
standard_normal_complex(size)	return (R + 1.j*I) for independent R and I from np.random.standard_normal.

Module description

Provide some random matrix ensembles for numpy.

The implemented ensembles are:

ensemble	matrix class drawn from	measure	invariant under	beta
GOE	real, symmetric	~ exp(-n/4 tr(H^2))	orthogonal O	1
GUE	hermitian	~ exp(-n/2 tr(H^2))	unitary U	2
CRE	O(n)	Haar	orthogonal O	/
COE	U in U(n) with U = U^T	Haar	orthogonal O	1
CUE	U(n)	Haar	unitary U	2
O_close_1	O(n)	?	/	/
U_close_1	U(n)	?	/	/

All functions in this module take a tuple (n, n) as first argument, such that we can use the function from_func() to generate a block diagonal Array with the block from the corresponding ensemble, for example:

npc.Array.from_func_square(GOE, [leg, leg.conj()])