Schemassupervised/v1

SuperConfig

Entry point

Language-independent config for an rtemis supervised-learning run. Mirrors the `SuperConfig` object: data references, optional preprocessing / decomposition, an algorithm with hyperparameters, optional tuning and outer resampling, and execution settings. The same config drives rtemis (R), rtemis CLI/shell, and rtemislive.

Raw JSONv1Unknown properties rejected

Properties

outcome: string | null
Name of the outcome column. Unset follows rtemis's convention, the last column.
features: array | null≥ 1 items
Names of the columns to use as predictors. Unset uses every column except the outcome.
weights: string | null
Name of the column to use as case weights.
positive_class: string | null
Outcome level to treat as positive (binary classification).
preprocessor_config: SupervisedPreprocessorConfig | null
decomposition_config: DecompositionConfig | null
hyperparameters: any
One configuration, searched as the product of its values, or a named set of them searched as their union. Write the single configuration unless the space to search is not a product.
tuner_config: TunerConfig | null
outer_resampling_config: ResamplerConfig | null
execution_config: ExecutionConfig
question: string | null
User-provided label / question for the run.
outdir: string | null
Output directory for results. Unset does not write to disk.
verbosity: integer≥ 0
Verbosity level.
dat_training_path: string | null
Path to the training data.
dat_validation_path: string | null
Path to the validation data.
dat_test_path: string | null
Path to the held-out test data.
character2factor: boolean
Read character columns as factors. Supervised learning needs categorical predictors as factors, and a delimited file cannot say which columns those are.

Relationships

Composes

decompositionexecutionhyperparametersresamplersupervisedpreprocessortuner