optunaz.utils package

Subpackages

• optunaz.utils.enums package
  • Submodules
  • optunaz.utils.enums.building_configuration_enum module
  • optunaz.utils.enums.configuration_enum module
  • optunaz.utils.enums.interface_enum module
  • optunaz.utils.enums.model_runner_enum module
  • optunaz.utils.enums.objective_enum module
  • optunaz.utils.enums.optimization_configuration_enum module
  • optunaz.utils.enums.prediction_configuration_enum module
  • optunaz.utils.enums.return_values_enum module
  • optunaz.utils.enums.visualization_enum module
  • Module contents
• optunaz.utils.preprocessing package
  • Submodules
  • optunaz.utils.preprocessing.deduplicator module
  • optunaz.utils.preprocessing.splitter module
  • optunaz.utils.preprocessing.transform module
  • Module contents

Submodules

optunaz.utils.files_paths module

optunaz.utils.files_paths.move_up_directory(path, n=1)

Function, to move up “n” directories for a given “path”.

optunaz.utils.files_paths.attach_root_path(path)

Function to attach the root path of the module for a given “path”.

optunaz.utils.load_json module

optunaz.utils.load_json.loadJSON(path)

optunaz.utils.mlflow module

optunaz.utils.mlflow.add_ellipsis(name, max_length=100)

optunaz.utils.mlflow.shorten_names(params)

class optunaz.utils.mlflow.MLflowCallback(trial_number_offset, tracking_uri=None, optconfig=None)

Bases: object

Callback to track Optuna trials with MLflow.

This callback adds Optuna-tracked information to MLflow. The MLflow experiment will be named after the Optuna study name. MLflow runs will be named after Optuna trial numbers.

trial_number_offset

tracking_uri = None

The URI of the MLflow tracking server. Please refer to mlflow.set_tracking_uri for more details.

optconfig = None

prepare_tags(study, trial)

Sets the tags for MLflow.

tmp_buildconfig(study, trial)

Creates a temporary file with build configuration, and returns a path to it.

optunaz.utils.retraining module

exception optunaz.utils.retraining.NoRetrainingDataConvention(task, message='input-directory file [{0}] does not contain a date format %Y-%m-%d')

Bases: Exception

Raised if a file in input-directory does not follow the %Y-%m-%d convention

exception optunaz.utils.retraining.NoNewRetrainingData

Bases: Exception

Raised if no new retraining data is available

exception optunaz.utils.retraining.NoDifferingRetrainingData

Bases: Exception

Raised if no different retraining data is available between previous & current time bins

exception optunaz.utils.retraining.RetrainingHeadersIssue

Bases: Exception

Raised when issue with retraining headers in a file (columns unknown)

exception optunaz.utils.retraining.RetrainingIsAlreadyProcessed(task, message='Retraining[{0}] already processed')

Bases: Exception

Raised when retraining is processed

exception optunaz.utils.retraining.RetrainingIsLocked(task, message='Retraining[{0}] is locked')

Bases: Exception

Raised when retraining is locked

exception optunaz.utils.retraining.TemporalPredsPredicted(task, message='Retraining[{0}] code is predicted')

Bases: Exception

Raised when a temporal prediction is already predicted.

exception optunaz.utils.retraining.NoPreviousModel(prev_model_name, message='No previous model found for [{0}]')

Bases: Exception

Raised when no previous model exists for an retraining

exception optunaz.utils.retraining.SamePreviousModel(task, message='Retraining[{0}] already processed')

Bases: Exception

Raised when a temporal prediction would be for the same (identical) model training

exception optunaz.utils.retraining.TimepointSkipped(task, message='Retraining[{0}] set to be skipped')

Bases: Exception

Raised when a timepoint should be skipped

exception optunaz.utils.retraining.SlurmNoLog

Bases: Exception

Raised when a SLURM job file is not present for submitted itcode jobs

exception optunaz.utils.retraining.SlurmTimeLimitExceeded

Bases: Exception

Raised when a past SLURM job time was exceeded

exception optunaz.utils.retraining.SlurmMemoryExceeded

Bases: Exception

Raised when a past SLURM memory was exceeded

exception optunaz.utils.retraining.SlurmJobSkip

Bases: Exception

Raised when an itcode is no longer trialed with Qptuna (i.e. due to incompatibility)

exception optunaz.utils.retraining.SlurmParseError

Bases: Exception

Raised when a .sh SLURM job modification (for resubmission) is not possible

optunaz.utils.schema module

optunaz.utils.schema.create_dependency(key, value)

optunaz.utils.schema.remove_schema_properties(schema, properties)

optunaz.utils.schema.add_boolean_guards_for_schema_properties(schema, properties)

optunaz.utils.schema.replacekey(input_)

optunaz.utils.schema.replacevalue(input_)

optunaz.utils.schema.addsibling(input_)

optunaz.utils.schema.delsibling(input_, siblings)

optunaz.utils.schema.getref(path, context)

Recursively returns nested items from a dict.

optunaz.utils.schema.copytitle(input_, context)

Copies “title” from “$ref” into oneOf.

optunaz.utils.schema.replaceenum(input_)

Replace singleton enums with const.

optunaz.utils.schema.addtitles(schema)

optunaz.utils.tracking module

optunaz.utils.tracking.get_authorization_header()

class optunaz.utils.tracking.TrackingData(trial_number, trial_value, scoring, trial_state, all_cv_test_scores, buildconfig)

Bases: object

Dataclass defining internal tracking format

trial_number

trial_value

scoring

trial_state

all_cv_test_scores

buildconfig

optunaz.utils.tracking.removeprefix(line, prefix)

optunaz.utils.tracking.round_scores(test_scores)

class optunaz.utils.tracking.InternalTrackingCallback(optconfig, trial_number_offset)

Bases: object

Callback to track (log) progress using internal tracking format

optconfig

trial_number_offset

class optunaz.utils.tracking.Datapoint(smiles: str, expected: float, predicted: float)

Bases: object

smiles

expected

predicted

class optunaz.utils.tracking.BuildTrackingData(response_column_name, test_scores, test_points)

Bases: object

Dataclass defining internal tracking format

response_column_name

test_scores

test_points

optunaz.utils.tracking.track_build(model, buildconfig)

Module contents

optunaz.utils.mkdict(obj)

optunaz.utils.load_df_from_file(filename, smiles_col)

optunaz.utils.remove_failed_idx(failed_idx, y_, smis, auxs)

optunaz.utils.md5_hash(dictionary)

MD5 hash of an optimisation algorithm or a model metadata dictionary