xplogger.experiment_manager.notebook package

Submodules

xplogger.experiment_manager.notebook.utils module

Utlities functions to make it easier to use xplogger with a jupyter notebook.

xplogger.experiment_manager.notebook.utils.get_mean_and_std_err(experiment_sequence: ExperimentSequence, metadata: DictConfig) → tuple[np.typing.NDArray[np.float32], np.typing.NDArray[np.float32], int]

Compute the mean and standard error for a given experiment sequence.

Parameters

• experiment_sequence (ExperimentSequence) –

• metadata (DictConfig) – metadata to use for computing the metrics

Returns

tuple of mean, standard error and number of experiments in the experiment sequence (useful for computing standard deviation etc).

Return type

tuple[np.typing.NDArray[np.float32], np.typing.NDArray[np.float32], int]

xplogger.experiment_manager.notebook.utils.make_df(metadata: DictConfig, step_metadata: DictConfig, groups: dict[Any, RecordList], hyperparams: dict[str, set[ValueType]], exp_seq_dict: ExperimentSequenceDict) → pd.DataFrame

Make a dataframe using the given experience sequence dict.

Parameters

• metadata (DictConfig) – Contains information like metric_name for the metrics of interest.

• step_metadata (DictConfig) – Contains information like metric_name for the step metric (eg epoch or frames)

• groups (dict[Any, RecordList]) –

• hyperparams (dict[str, set[ValueType]]) –

• exp_seq_dict (ExperimentSequenceDict) –

Returns

Return type

pd.DataFrame

xplogger.experiment_manager.notebook.utils.prettyprint_dict(d: dict, sep: str = '\t', indent: int = 0) → None

Pretty print a dictionary.

Parameters

• d (dict) – input dictionary

• sep (str, optional) – Seperator to use. Defaults to “t”.

• indent (int, optional) – Indentation to use. Defaults to 0.

Module contents