hex

Class Model.Parameters

java.lang.Object

water.Iced

hex.Model.Parameters

All Implemented Interfaces:

java.io.Externalizable, java.io.Serializable, java.lang.Cloneable, Freezable

Direct Known Subclasses:

ClusteringModel.ClusteringParameters, QuantileModel.QuantileParameters

Enclosing class:

Model<M extends Model<M,P,O>,P extends Model.Parameters,O extends Model.Output>

public abstract static class Model.Parameters
extends Iced

Model-specific parameter class. Each model sub-class contains an instance of one of these containing its builder parameters, with model-specific parameters. E.g. KMeansModel extends Model and has a KMeansParameters extending Model.Parameters; sample parameters include K, whether or not to normalize, max iterations and the initial random seed.

The non-transient fields are input parameters to the model-building process, and are considered "first class citizens" by the front-end - the front-end will cache Parameters (in the browser, in JavaScript, on disk) and rebuild Parameter instances from those caches.

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
boolean	_balance_classes Should all classes be over/under-sampled to balance the class distribution?
float[]	_class_sampling_factors Desired over/under-sampling ratios per class (lexicographic order).
boolean	_ignore_const_cols
java.lang.String[]	_ignored_columns
float	_max_after_balance_size When classes are being balanced, limit the resulting dataset size to the specified multiple of the original dataset size.
int	_max_confusion_matrix_size For classification models, the maximum size (in terms of classes) of the confusion matrix for it to be printed.
int	_max_hit_ratio_k The maximum number (top K) of predictions to use for hit ratio computation (for multi-class only, 0 to disable)
Key<Frame>	_model_id
java.lang.String	_offset_column
java.lang.String	_response_column Supervised models have an expected response they get to train with!
boolean	_score_each_iteration
Key<Frame>	_train
Key<Frame>	_valid
java.lang.String	_weights_column

Constructor Summary

Constructors

Constructor and Description
Model.Parameters()

Method Summary

Methods

Modifier and Type	Method and Description
protected long	checksum_impl() Compute a checksum based on all non-transient non-static ice-able assignable fields (incl.
protected boolean	defaultDropConsCols()
protected boolean	defaultDropNA20Cols()
double	missingColumnsType() Type of missing columns during adaptation between train/test datasets Overload this method for models that have sparse data handling - a zero will preserve the sparseness.
void	read_lock_frames(Job job) Read-Lock both training and validation User frames.
void	read_unlock_frames(Job job) Read-UnLock both training and validation User frames.
Frame	train()
Frame	valid()

Methods inherited from class water.Iced

clone, frozenType, read_impl, read, readExternal, readJSON_impl, readJSON, toJsonString, write_impl, write, writeExternal, writeHTML_impl, writeHTML, writeJSON_impl, writeJSON

Methods inherited from class java.lang.Object

equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

_model_id

public Key<Frame> _model_id

_train

public Key<Frame> _train

_valid

public Key<Frame> _valid

_ignored_columns

public java.lang.String[] _ignored_columns

_ignore_const_cols

public boolean _ignore_const_cols

_weights_column

public java.lang.String _weights_column

_offset_column

public java.lang.String _offset_column

_score_each_iteration

public boolean _score_each_iteration

_response_column

public java.lang.String _response_column

Supervised models have an expected response they get to train with!

_balance_classes

public boolean _balance_classes

Should all classes be over/under-sampled to balance the class distribution?

_max_after_balance_size

public float _max_after_balance_size

When classes are being balanced, limit the resulting dataset size to the specified multiple of the original dataset size. Maximum relative size of the training data after balancing class counts (can be less than 1.0)

_class_sampling_factors

public float[] _class_sampling_factors

Desired over/under-sampling ratios per class (lexicographic order). Only when balance_classes is enabled. If not specified, they will be automatically computed to obtain class balance during training.

_max_hit_ratio_k

public int _max_hit_ratio_k

The maximum number (top K) of predictions to use for hit ratio computation (for multi-class only, 0 to disable)

_max_confusion_matrix_size

public int _max_confusion_matrix_size

For classification models, the maximum size (in terms of classes) of the confusion matrix for it to be printed. This option is meant to avoid printing extremely large confusion matrices.

Constructor Detail

Model.Parameters

public Model.Parameters()

Method Detail

train

public final Frame train()

Returns:

the training frame instance

valid

public final Frame valid()

Returns:

the validation frame instance, or null if a validation frame was not specified

read_lock_frames

public void read_lock_frames(Job job)

Read-Lock both training and validation User frames.

read_unlock_frames

public void read_unlock_frames(Job job)

Read-UnLock both training and validation User frames.

defaultDropNA20Cols

protected boolean defaultDropNA20Cols()

defaultDropConsCols

protected boolean defaultDropConsCols()

missingColumnsType

public double missingColumnsType()

Type of missing columns during adaptation between train/test datasets Overload this method for models that have sparse data handling - a zero will preserve the sparseness. Otherwise, NaN is used.

Returns:

real-valued number (can be NaN)

checksum_impl

protected long checksum_impl()

Compute a checksum based on all non-transient non-static ice-able assignable fields (incl. inherited ones) which have @API annotations. Sort the fields first, since reflection gives us the fields in random order and we don't want the checksum to be affected by the field order. NOTE: if a field is added to a Parameters class the checksum will differ even when all the previous parameters have the same value. If a client wants backward compatibility they will need to compare parameter values explicitly.

Returns:

checksum