- class ray.tune.ExperimentAnalysis(experiment_checkpoint_path: str | PathLike, *, storage_filesystem: pyarrow.fs.FileSystem | None = None, trials: List[Trial] | None = None, default_metric: str | None = None, default_mode: str | None = None)#
Analyze results from a Ray Train/Tune experiment.
To use this class, the run must store the history of reported metrics in log files (e.g.,
progress.csv). This is the default behavior, unless default loggers are explicitly excluded with the
experiment_checkpoint_path – Path to an
experiment_state.jsonfile, or a directory that contains an
default_metric – Default metric for comparing results. Can be overwritten with the
metricparameter in the respective functions.
default_mode – Default mode for comparing results. Has to be one of [min, max]. Can be overwritten with the
modeparameter in the respective functions.
trials – List of trials that can be accessed via
PublicAPI (beta): This API is in beta and may change before becoming stable.
Returns a pandas.DataFrame object constructed from the trials.
Returns all trial hyperparameter configurations.
Gets best persistent checkpoint path of provided trial.
Retrieve the best config corresponding to the trial.
Retrieve the best trial object.
Gets the last checkpoint of the provided trial, i.e., with the highest "training_iteration".
Get the checkpoint path of the best trial of the experiment
Get the config of the best trial of the experiment
Get the full result dataframe of the best trial of the experiment
Get the last result of the best trial of the experiment
Get the best result of the experiment as a pandas dataframe.
Get the best trial of the experiment
Path pointing to the experiment directory on persistent storage.
Get the last result of the all trials of the experiment
Get all the last results as a pandas dataframe.
List of all dataframes of the trials.