Astronomical Visibility Components

class tabascal.components.ast_vis.GPVisAst

build_constants()

Return arrays that do not change during the forward pass.

Returns a dict of array_name -> array_value. These will be stored in constants as “_c/<ClassName>/array_name” by Model.__init__.

build_forward()

Return pure, JIT-compatible function

build_set_params()

Build parameter sampling function (optional)

config_params: Dict[str, Param] = {'ast.freq_pad_factor': Param(types=(<class 'int'>, <class 'float'>), default=2, choices=(), item=None, gt=None, ge=1, lt=None, le=None, null_ok=False, doc='padding of the modelled frequency interval, to avoid periodicity'), 'ast.init': Param(types=(), default='sample', choices=('data', 'prior', 'truth', 'sample'), item=None, gt=None, ge=None, lt=None, le=None, null_ok=False, doc='how the astronomical visibility parameters are initialised'), 'ast.mean': Param(types=(), default=0, choices=('data', 'zeros', 0), item=None, gt=None, ge=None, lt=None, le=None, null_ok=False, doc='mean of the prior over the astronomical visibilities'), 'ast.pow_spec.cutoff': Param(types=(<class 'int'>, <class 'float'>), default=1e-06, choices=(), item=None, gt=0, ge=None, lt=None, le=None, null_ok=False, doc='relative power below which a Fourier component is not modelled'), 'ast.pow_spec.fov_deg': Param(types=(<class 'int'>, <class 'float'>), default=<derived from the data>, choices=(), item=None, gt=0, ge=None, lt=None, le=None, null_ok=False, doc='full field of view in degrees; null uses the MS dish diameter'), 'ast.pow_spec.gammas': Param(types=(<class 'list'>,), default=<required>, choices=(), item=(<class 'int'>, <class 'float'>), gt=None, ge=None, lt=None, le=None, null_ok=False, doc='power-law drop-off per axis, [freq, time]'), 'ast.pow_spec.k0_freq': Param(types=(<class 'int'>, <class 'float'>), default=<required>, choices=(), item=None, gt=0, ge=None, lt=None, le=None, null_ok=False, doc='inverse correlation scale along the frequency axis'), 'ast.pow_spec.p0': Param(types=(<class 'int'>, <class 'float'>), default=<required>, choices=(), item=None, gt=0, ge=None, lt=None, le=None, null_ok=False, doc='mean power of the astronomical signal'), 'ast.time_pad_factor': Param(types=(<class 'int'>, <class 'float'>), default=2, choices=(), item=None, gt=None, ge=1, lt=None, le=None, null_ok=False, doc='padding of the modelled time interval, to avoid periodicity')}

Config parameters this component reads, keyed by their dotted path in the config file. Declared here, next to the code that reads them, so a component and its requirements can never drift apart. Only the components named in model.components contribute to a run’s schema – see tabascal.config_schema.collect_params(). Subclasses extend (and may override) what their base declared; the parameters read outside any component live on tabascal.config.TabConfig.

setup(config)

All validation and error-prone operations here

validate_and_test()

Call this before using in JIT context