cellflow.training.Metrics.on_log_iteration

Metrics.on_log_iteration(valid_source_data, valid_true_data, valid_pred_data, solver)

Called at each validation/log iteration to compute metrics

Parameters:

• valid_source_data (dict[str, dict[str, ndarray[tuple[Any, ...], dtype[float64]]]]) – Source data in nested dictionary format with same keys as valid_true_data

• valid_true_data (dict[str, dict[str, ndarray[tuple[Any, ...], dtype[float64]]]]) – Validation data in nested dictionary format with same keys as valid_pred_data

• valid_pred_data (dict[str, dict[str, ndarray[tuple[Any, ...], dtype[float64]]]]) – Predicted data in nested dictionary format with same keys as valid_true_data

• solver (OTFlowMatching | GENOT) – OTFlowMatching solver or GENOT solver with a conditional velocity field.

Return type:

dict[str, float]

Returns:

: Computed metrics between the true validation data and predicted validation data as a dictionary