modularml.core.splitters

class modularml.core.splitters.BaseSplitter

Bases: ABC

abstractmethod split(samples: List[Sample]) → Dict[str, List[str]]

Returns a dictionary mapping subset names to Sample.uuid.

class modularml.core.splitters.ConditionSplitter(**conditions: Dict[str, Dict[str, Any | List[Any] | Callable]])

Bases: BaseSplitter

Splits samples into subsets based on user-defined filtering conditions.

Parameters:

conditions (Dict[str, Dict[str, Union[Any, List[Any], Callable]]]) – The outer dict defines keys representing the new subset names. Within each subset, an inner dict defines the filtering conditions to construct the subset. The inner dictionary uses that same format as the FeatureSet.filter()) method.

Examples: Below defines three subsets (‘low_temp’, ‘high_temp’, and ‘cell_5’). The ‘low_temp’ subset contains all samples with temperatures under 20, the ‘high_temp’ subsets contains all samples with temperature greater than 20, and the ‘cell_5’ subset contains all samples where cell_id is 5. Note that subsets can have overlapping samples if the split conditions are not carefully **defined. A UserWarning will be raised when this happens, **

``` python

ConditionSplitter(

low_temp={‘temperature’: lambda x: x < 20}, high_temp={‘temperature’: lambda x: x >= 20}, cell_5={‘cell_id’: 5}

)

```

__init__(**conditions: Dict[str, Dict[str, Any | List[Any] | Callable]])

Splits samples into subsets based on user-defined filtering conditions.

Parameters:

conditions (Dict[str, Dict[str, Union[Any, List[Any], Callable]]]) – The outer dict defines keys representing the new subset names. Within each subset, an inner dict defines the filtering conditions to construct the subset. The inner dictionary uses that same format as the FeatureSet.filter()) method.

Examples: Below defines three subsets (‘low_temp’, ‘high_temp’, and ‘cell_5’). The ‘low_temp’ subset contains all samples with temperatures under 20, the ‘high_temp’ subsets contains all samples with temperature greater than 20, and the ‘cell_5’ subset contains all samples where cell_id is 5. Note that subsets can have overlapping samples if the split conditions are not carefully **defined. A UserWarning will be raised when this happens, **

``` python

ConditionSplitter(

low_temp={‘temperature’: lambda x: x < 20}, high_temp={‘temperature’: lambda x: x >= 20}, cell_5={‘cell_id’: 5}

)

```

split(samples: List[Sample]) → Dict[str, List[str]]

Applies the condition-based split on the given samples.

Parameters:

samples (List[Sample]) – The list of samples to split.

Returns:

Dictionary mapping subset names to Sample.uuid.

Return type:

Dict[str, List[str]]

class modularml.core.splitters.RandomSplitter(ratios: Dict[str, float], seed: int = 42)

Bases: BaseSplitter

Creates a random splitter based on sample ratios

Parameters:

• ratios (Dict[str, float]) – Keyword-arguments that define subset names and percent splits. E.g., RandomSplitter(train=0.5, test=0.5). All values must add to exactly 1.0.

• seed (int) – The seed of the random generator.

__init__(ratios: Dict[str, float], seed: int = 42)

Creates a random splitter based on sample ratios

Parameters:

• ratios (Dict[str, float]) – Keyword-arguments that define subset names and percent splits. E.g., RandomSplitter(train=0.5, test=0.5). All values must add to exactly 1.0.

• seed (int) – The seed of the random generator.

split(samples: List[Sample]) → Dict[str, List[str]]

Randomly splits a list of samples based on the defined ratios.

Parameters:

samples (List[Sample]) – The list of samples to split.

Returns:

Dictionary mapping subset names to Sample.uuid.

Return type:

Dict[str, List[str]]

Modules

conditon_splitter
random_splitter
splitter