conmo.experiment.Pipeline
- class conmo.experiment.Pipeline(dataset: Dataset, splitter: Optional[Splitter], preprocesses: Optional[Iterable[Preprocess]], algorithms: Iterable[Algorithm], metrics: Iterable[Metric])[source]
- __init__(dataset: Dataset, splitter: Optional[Splitter], preprocesses: Optional[Iterable[Preprocess]], algorithms: Iterable[Algorithm], metrics: Iterable[Metric]) None [source]
- generate_dirs(pipe_dir: str) None [source]
Auxiliary method to generate directories for each of the steps in the current pipeline.
- Parameters
pipe_dir (str) – Name of the pipe directory.
- run(pipe_dir: str, pipe_num: int, pipes: int) None [source]
Contains all the logic for the execution of a particular pipeline, creating intermediate directories for data passing and executing the relevant methods for each step.
- Parameters
pipe_dir (str) – Name of the current pipeline directory.
pipe_num (int) – Index of the current pipelines.
pipes (int) – Total number of pipelines in the current experiment.
Methods
__init__
(dataset, splitter, preprocesses, ...)generate_dirs
(pipe_dir)Auxiliary method to generate directories for each of the steps in the current pipeline.
run
(pipe_dir, pipe_num, pipes)Contains all the logic for the execution of a particular pipeline, creating intermediate directories for data passing and executing the relevant methods for each step.
Attributes
ALGORITHMS_FOLDER
DATASET_FOLDER
METRICS_FOLDER
PREPROCESSES_FOLDER
SPLITTER_FOLDER