Configuration

Sionna PHY’s configuration API. It can be used to set global variables which can be usedby all of its modules and functions.

classsionna.phy.config.Config[source]

Sionna PHY Configuration Class

This singleton class is used to define global configuration variablesand random number generators that can be accessed from all modulesand functions. It is instantiated immediately and its properties can beaccessed assionna.phy.config.desired_property.

propertynp_cdtype

Default NumPy dtype for complex floating point numbers

Type:

np.dtype

propertynp_rdtype

Default NumPy dtype for real floating point numbers

Type:

np.dtype

propertynp_rng

NumPy random number generator

fromsionna.phyimportconfigconfig.seed=42# Set seed for deterministic results# Use generator instead of np.randomnoise=config.np_rng.normal(size=[4])
Type:

np.random.Generator

propertyprecision

Default precision used for all computations

The “single” option represents real-valued floating-point numbersusing 32 bits, whereas the “double” option uses 64 bits.For complex-valued data types, each component of the complex number(real and imaginary parts) uses either 32 bits (for “single”)or 64 bits (for “double”).

Type:

“single” (default) | “double”

propertypy_rng

Python random number generator

fromsionna.phyimportconfigconfig.seed=42# Set seed for deterministic results# Use generator instead of randomint=config.py_rng.randint(0,10)
Type:

random.Random

propertyseed

Get/set seed for all random number generators

All random number generators used internally by Sionnacan be configured with a common seed to ensure reproducabilityof results. It defaults toNone which implies that a randomseed will be used and results are non-deterministic.

# This code will lead to deterministic resultsfromsionna.phyimportconfigfromsionna.phy.mappingimportBinarySourceconfig.seed=42print(BinarySource()([10]))
tf.Tensor([0. 1. 1. 1. 1. 0. 1. 0. 1. 0.], shape=(10,), dtype=float32)
Type:

None (default) |int

propertytf_cdtype

Default TensorFlow dtype for complex floating point numbers

Type:

tf.dtype

propertytf_rdtype

Default TensorFlow dtype for real floating point numbers

Type:

tf.dtype

propertytf_rng

TensorFlow random number generator

fromsionna.phyimportconfigconfig.seed=42# Set seed for deterministic results# Use generator instead of tf.randomnoise=config.tf_rng.normal([4])
Type:

tf.random.Generator