skbase.validate.check_sequence#

skbase.validate.check_sequence(input_seq: Sequence[Any], sequence_type: type | Tuple[type, ...] | None = None, element_type: type | Tuple[type, ...] | None = None, coerce_output_type_to: type | None = None, coerce_scalar_input: bool = False, sequence_name: str | None = None) → Sequence[Any][source]#

Check whether an object is a sequence with optional check of element types.

If element_type is supplied all elements are also checked against provided types.

Parameters:

input_seqAny

The input sequence to be validated.

sequence_typetype or tuple[type], default=None

The allowed sequence type that seq can be an instance of.

element_typetype or tuple[type], default=None

The allowed type(s) for elements of seq.

coerce_output_type_tosequence type

The sequence type that the output sequence should be coerced to.

If None, then the output sequence is the same as input sequence.
If a sequence type (e.g., list, tuple) is provided then the output sequence is coerced to that type.

coerce_scalar_inputbool, default=False

Whether scalar input should be coerced to a sequence type prior to running the check. If True, a scalar input like will be coerced to a tuple containing a single scalar. To output a sequence type other than a tuple, set the coerce_output_type_to keyword to the desired sequence type (e.g., list).

sequence_namestr, default=None

Name of input_seq to use if error messages are raised.

Returns:

Sequence: The input sequence if has expected type.

Raises:

TypeError: If seq is not instance of sequence_type or element_type is not None and all elements are not instances of element_type.

Examples

>>> from skbase.base import BaseEstimator, BaseObject
>>> from skbase.validate import is_sequence

>>> check_sequence([1, 2, 3])
[1, 2, 3]

Generators are not sequences so an error is raised

>>> check_sequence((c for c in [1, 2, 3])) 

The check can require a certain type of sequence

>>> check_sequence([1, 2, 3, 4], sequence_type=list)
[1, 2, 3, 4]

Expected to raise and error because the input is not a tuple

>>> check_sequence([1, 2, 3, 4], sequence_type=tuple) 

It is also possible to check the type of sequence elements

>>> check_sequence([1, 2, 3], element_type=int)
[1, 2, 3]
>>> check_sequence([1, 2, 3, 4], sequence_type=list, element_type=(int, float))
[1, 2, 3, 4]

The check also works with BaseObjects

>>> check_sequence((BaseObject(), BaseEstimator()), element_type=BaseObject)
(BaseObject(), BaseEstimator())