treeflow.model.ml module

treeflow.model.ml.fit_fixed_topology_maximum_likelihood_sgd(model: Distribution, topologies: Dict[str, TensorflowTreeTopology], num_steps: int = 10000, optimizer: Optimizer = None, trace_fn: Callable[[MinimizeTraceableQuantities], object] | None = None, convergence_criterion: ConvergenceCriterion | None = None, init: object | None = None, return_full_length_trace: bool = False, dtype=tf.float64, transform_traced_params: bool = True, progress_bar: bool | ProgressBarFunc = False, progress_bar_step: int = 10, **opt_kwargs)

Returns:

• variables – Maximum likelihood estimate for model variables

• trace – Optimization trace

• bijector – Bijector to transform unconstrained parameters to model variables