lale.lib.rasl.project module¶
- class lale.lib.rasl.project.Project(*, columns=None, drop_columns=None)¶
Bases:
PlannedIndividualOp
Projection keeps a subset of the columns, like in relational algebra.
This documentation is auto-generated from JSON schemas.
Examples
>>> df = pd.DataFrame(data={'A': [1,2], 'B': ['x','y'], 'C': [3,4]}) >>> keep_numbers = Project(columns={'type': 'number'}) >>> keep_numbers.fit(df).transform(df) NDArrayWithSchema([[1, 3], [2, 4]])
- Parameters
columns (union type, not for optimizer, default None) –
The subset of columns to retain.
The supported column specification formats include some of the ones from scikit-learn’s ColumnTransformer, and in addition, filtering by using a JSON subschema check.
None
If not specified, keep all columns.
or array of items : integer
Multiple columns by index.
or array of items : string
Multiple Dataframe columns by names.
or callable
Callable that is passed the input data X and can return a list of column names or indices.
or dict
Keep columns whose schema is a subschema of this JSON schema.
drop_columns (union type, not for optimizer, default None) –
The subset of columns to remove.
The drop_columns argument supports the same formats as columns. If both are specified, keep everything from columns that is not also in drop_columns.
None
If not specified, drop no further columns.
or array of items : integer
Multiple columns by index.
or array of items : string
Multiple Dataframe columns by names.
or callable
Callable that is passed the input data X and can return a list of column names or indices.
or dict
Remove columns whose schema is a subschema of this JSON schema.
- fit(X, y=None, **fit_params)¶
Train the operator.
Note: The fit method is not available until this operator is trainable.
Once this method is available, it will have the following signature:
- Parameters
X (array) –
Features; the outer array is over samples.
items : array
items : union type
float
or string
y (any type, optional) – Target for supervised learning (ignored).
- partial_fit(X, y=None, **fit_params)¶
Incremental fit to train train the operator on a batch of samples.
Note: The partial_fit method is not available until this operator is trainable.
Once this method is available, it will have the following signature:
- transform(X, y=None)¶
Transform the data.
Note: The transform method is not available until this operator is trained.
Once this method is available, it will have the following signature:
- Parameters
X (array) –
Features; the outer array is over samples.
items : array
items : union type
float
or string
- Returns
result – Features; the outer array is over samples.
items : array
items : union type
float
or string
- Return type
array
- lale.lib.rasl.project.get_column_factory(columns, kind) MonoidFactory [source]¶