ParamGridBuilder

class pyspark.ml.tuning.ParamGridBuilder

Builder for a param grid used in grid search-based model selection.

New in version 1.4.0.

Examples

>>> from pyspark.ml.classification import LogisticRegression
>>> lr = LogisticRegression()
>>> output = ParamGridBuilder() \
...     .baseOn({lr.labelCol: 'l'}) \
...     .baseOn([lr.predictionCol, 'p']) \
...     .addGrid(lr.regParam, [1.0, 2.0]) \
...     .addGrid(lr.maxIter, [1, 5]) \
...     .build()
>>> expected = [
...     {lr.regParam: 1.0, lr.maxIter: 1, lr.labelCol: 'l', lr.predictionCol: 'p'},
...     {lr.regParam: 2.0, lr.maxIter: 1, lr.labelCol: 'l', lr.predictionCol: 'p'},
...     {lr.regParam: 1.0, lr.maxIter: 5, lr.labelCol: 'l', lr.predictionCol: 'p'},
...     {lr.regParam: 2.0, lr.maxIter: 5, lr.labelCol: 'l', lr.predictionCol: 'p'}]
>>> len(output) == len(expected)
True
>>> all([m in expected for m in output])
True

Methods

addGrid(param, values)	Sets the given parameters in this grid to fixed values.
baseOn(*args)	Sets the given parameters in this grid to fixed values.
build()	Builds and returns all combinations of parameters specified by the param grid.

Methods Documentation

addGrid(param: pyspark.ml.param.Param[Any], values: List[Any]) → pyspark.ml.tuning.ParamGridBuilder

Sets the given parameters in this grid to fixed values.

param must be an instance of Param associated with an instance of Params (such as Estimator or Transformer).

New in version 1.4.0.

baseOn(*args: Union[ParamMap, Tuple[pyspark.ml.param.Param, Any]]) → ParamGridBuilder

Sets the given parameters in this grid to fixed values. Accepts either a parameter dictionary or a list of (parameter, value) pairs.

New in version 1.4.0.

build() → List[ParamMap]

Builds and returns all combinations of parameters specified by the param grid.

New in version 1.4.0.

SummaryBuilder CrossValidator