`palma.utils.checker`#

Module Contents#

ProjectPlanChecker is an object that checks the project plan.

`_CLASSIFICATION_METRICS`
`_REGRESSION_METRICS`

palma.utils.checker._CLASSIFICATION_METRICS = ['accuracy', 'balanced_accuracy', 'top_k_accuracy', 'average_precision', 'neg_brier_score',...#

palma.utils.checker._REGRESSION_METRICS = ['explained_variance', 'r2', 'max_error', 'neg_median_absolute_error',...#

class palma.utils.checker.ProjectPlanChecker#

Bases: object

ProjectPlanChecker is an object that checks the project plan.

At the build() moment, this object run several checks in order to see if the project plan is well designed.

Here is an overview of the checks performed by the object:

_check_arrays() : see whether X and y attribute are compliant with sklearn standards.
_check_project_problem(): see if the problem type is correctly informed by the user.
_check_problem_metrics(): see if the known metrics are consistent with the project problem

run_checks(project: palma.Project) → None#

Perform some tests on the project plan

Several checks are performed in order to check if the project plan is consistent:

checks the project problem

checks the metrics provided by the user

checks the data provided by the user (scikit learn wrapper)

Parameters: