bgtrees package

Subpackages

• bgtrees.finite_gpufields package
  • Subpackages
    • bgtrees.finite_gpufields.cuda_operators package
      • Submodules
      • bgtrees.finite_gpufields.cuda_operators.check_inverse module
      • bgtrees.finite_gpufields.cuda_operators.dot_product module
      • bgtrees.finite_gpufields.cuda_operators.inverse module
      • bgtrees.finite_gpufields.cuda_operators.py_dotproduct module
      • Module contents
  • Submodules
  • bgtrees.finite_gpufields.finite_fields_tf module
    • FiniteField
      • FiniteField.Spec
      • FiniteField.n
      • FiniteField.p
      • FiniteField.reshape_ff()
      • FiniteField.shape
      • FiniteField.transpose_ff()
      • FiniteField.values
      • FiniteField.zero_panning()
    • finite_field_expand_dims()
    • finite_field_reduce_prod()
    • finite_field_reduce_sum()
    • finite_field_squeeze()
    • finite_field_unary_operations()
    • finite_field_unstack()
    • generate_finite_field()
    • get_imaginary_for()
    • test_me()
  • bgtrees.finite_gpufields.operations module
    • ff_dot_product()
    • ff_dot_product_single_batch()
    • ff_dot_product_tris()
    • ff_dot_product_tris_single_batch()
    • ff_einsum_generic()
    • ff_tensor_product()
  • Module contents

Submodules

bgtrees.currents module

bgtrees.currents.J_μ(lmoms, lpols, put_propagator=True, depth=0, verbose=False, einsum=<function einsum>)

Recursive vectorized current builder. End of recursion is polarization tensors.

TODO: try to merge this and another_j

bgtrees.currents.another_j(lmoms, lpols, put_propagator=True, verbose=False)

Compute the current for an input array of shape

(replicas, multiplicity, dimension)

bgtrees.currents.forest(lmoms, lpols, verbose=False, einsum=<function einsum>)

bgtrees.metric_and_verticies module

bgtrees.metric_and_verticies.MinkowskiMetric(D)

D-dimensional Minkowski metric in the mostly negative convention.

bgtrees.metric_and_verticies.V3g(lp1, lp2, einsum=<function einsum>)

3-gluon vertex, upper indices μνρ, D-dimensional

bgtrees.metric_and_verticies.V4g(D)

4-gluon vertex, upper indices μνρσ

bgtrees.metric_and_verticies.diag_mink(D)

bgtrees.metric_and_verticies.new_V3g(lp1, lp2)

3-gluon vertex, upper indices μνρ, D-dimensional. Reduce tensor products.

bgtrees.metric_and_verticies.η(D)

D-dimensional Minkowski metric in the mostly negative convention.

bgtrees.settings module

class bgtrees.settings.Settings(use_gpu: bool = False, D: int = 4, dtype: type = <class 'numpy.int64'>, p: int = 2147483629)

Bases: object

D: int = 4

dtype

alias of int64

executing_eagerly()

p: int = 2147483629

run_tf_eagerly()

property tf_p

use_gpu: bool = False

bgtrees.states module

bgtrees.states.momflat(momD, momχ)

Massless (flat) projection of 4D massive momentum onto a reference direction momχ.

bgtrees.states.ε1(oParticles, index, einsum=<function einsum>)

Negative polarization vector: ε⁻^μ = <i|γ^μ|q]/(√2[iq])

bgtrees.states.ε2(oParticles, index, einsum=<function einsum>)

Positive polarization vector: ε⁺^μ = <q|γ^μ|i]/(√2<qi>)

bgtrees.states.εm(oParticles, index, einsum=<function einsum>)

Negative polarization vector: ε⁻^μ = <i|γ^μ|q]/(√2[iq])

bgtrees.states.εp(oParticles, index, einsum=<function einsum>)

Positive polarization vector: ε⁺^μ = <q|γ^μ|i]/(√2<qi>)

bgtrees.states.μ2(momD, d=None)

The 4D mass of a massless D momentum (d=None). For μ^2_d = k^2_4 + … + k^2_d-1

bgtrees.tools module

bgtrees.tools.gpu_constant(func)

Turns constants into gpu constants if needed.

bgtrees.tools.gpu_function(func)

Passes additional arguments to run on gpu if needed. Dispatchs a different function depending on whether the arguments are finite fields or not

bgtrees.tools.tsr(x)

Module contents