export

The module of exporting the experiment data. (qurry.qurrium.container.export)

class qurry.qurrium.experiment.export.Export(exp_id: str, exp_name: str, serial: int | None, summoner_id: str | None, summoner_name: str | None, filename: str, files: dict[str, str], args: dict[str, Any], commons: CommonparamsDict, outfields: dict[str, Any], adventures: dict[str, Any], legacy: dict[str, Any], tales: dict[str, Any], reports: dict[Hashable, dict[str, Any]], tales_reports: dict[str, dict[Hashable, dict[str, Any]]])

Data-stored namedtuple with all experiments data which is jsonable.

adventures: dict[str, Any]

Recording the data of ‘beforeward’, which will be packed into .advent.json. ~ A Great Adventure begins ~

args: dict[str, Any]

Construct the experiment’s parameters, which will be packed into .args.json.

commons: CommonparamsDict

Construct the experiment’s common parameters, which will be packed into .args.json.

exp_id: str

ID of experiment, which will be packed into .args.json.

exp_name: str

Name of the experiment, which will be packed into .args.json. If this experiment is called by multimanager, then this name will never apply as filename.

filename: str

The name of file to be exported, it will be decided by the export() when it’s called. More info in the pydoc of :prop:`files` or export(), which will be packed into .args.json.

files: dict[str, str]

The list of file to be exported.

### Single experiment:

For the write() function actually exports 4 different files respecting to adventure, legacy, tales, and reports like:

which bla_exp is the example filename.

### Multi-experiment:

If this experiment is called by :cls:`MultiManager`, then the it will be named after summoner_name as known as the name of :cls:`MultiManager`.

which BLBLA_project is the example :cls:`MultiManager` name stored at :prop:`commonparams.summoner_name`. At this senerio, the :prop:`exp_name` will never apply as filename.

legacy: dict[str, Any]

Recording the data of ‘afterward’, which will be packed into .legacy.json. ~ The Legacy remains from the achievement of ancestors ~

outfields: dict[str, Any]

Recording the data of other unused arguments, which will be packed into .args.json.

reports: dict[Hashable, dict[str, Any]]

Recording the data of ‘reports’, which will be packed into .reports.json.

### Reports format:

~ The guild concludes the results. ~

serial: int | None

Index of experiment in a multiOutput, which will be packed into .args.json.

summoner_id: str | None

ID of experiment of the multiManager, which will be packed into .args.json.

summoner_name: str | None

Name of experiment of the multiManager, which will be packed into .args.json.

tales: dict[str, Any]

Recording the data of ‘side_product’ in ‘afterward’ and ‘beforewards’ for API, which will be packed into .*.tales.json. ~ Tales of braves circulate ~

tales_reports: dict[str, dict[Hashable, dict[str, Any]]]

Recording the data of ‘side_product’ in ‘reports’ for API, which will be packed into .*.reprts.json.

### Tales Reports format:

~ Tales of braves circulate ~

write(multiprocess: bool = False, pbar: tqdm | None = None) → tuple[str, dict[str, str]]

Export the experiment data, if there is a previous export, then will overwrite.

Hint

This function will traversal all objects in the export_set, so it will ensure the jsonable of all objects. And this will reduce the performance of exporting.

Parameters:

• multiprocess (bool, optional) – Whether to use multiprocess to export, Defaults to False. It’s dangerous to use multiprocess to export. It may cause memory leak.

• pbar (Optional[tqdm.tqdm], optional) – The progress bar for exporting. Defaults to None.

Returns:

The first element is the id of experiment, the second element is the dictionary of files of experiment.

Return type:

tuple[str, dict[str, str]]