Package edu.cmu.tetrad.data

Class BoxDataSet

java.lang.Object

edu.cmu.tetrad.data.BoxDataSet

All Implemented Interfaces:

DataModel, DataSet, KnowledgeTransferable, VariableSource, TetradSerializable, Serializable

public final class BoxDataSet
extends Object
implements DataSet

Wraps a DataBox in such a way that mixed data sets can be stored. The type of each column must be specified by a Variable object, which must be either a ContinuousVariable or a DiscreteVariable. This class violates object orientation in that the underlying data matrix is retrievable using the getDoubleData() method. This is allowed so that external calculations may be performed on large datasets without having to allocate extra memory. If this matrix needs to be modified externally, please consider making a copy of it first, using the TetradMatrix copy() method.

The data set may be given a name; this name is not used internally.

The data set has a list of variables associated with it, as described above. This list is coordinated with the stored data, in that data for the i'th variable will be in the i'th column.

A subset of variables in the data set may be designated as selected. This selection set is stored with the data set and may be manipulated using the select and deselect methods.

A multiplicity m_i may be associated with each case c_i in the dataset, which is interpreted to mean that that c_i occurs m_i times in the dataset.

Knowledge may be associated with the data set, using the setKnowledge method. This knowledge is not used internally to the data set, but it may be retrieved by algorithms and used.

Version:

$Id: $Id

Author:

josephramsey

See Also:

• Variable
• Knowledge
• Serialized Form

Constructor Summary

Constructors

Constructor	Description
BoxDataSet(BoxDataSet dataSet)	Makes of copy of the given data set.
BoxDataSet(DataBox dataBox, List<Node> variables)	Constructs a new data set with the given number of rows and columns, with all values set to missing.

Method Summary

Modifier and Type	Method	Description
void	addVariable(int index, Node variable)	Adds the given variable at the given index.
void	addVariable(Node variable)	Adds the given variable to the data set.
void	changeVariable(Node from, Node to)	Changes the variable for the given column from from to to.
void	clearSelection()	Marks all variables as deselected.
DataSet	copy()	Returns a copy of this dataset.
void	ensureColumns(int columns, List<String> excludedVariableNames)	Ensures that the dataset has at least columns columns.
void	ensureRows(int rows)	Ensures that the dataset has at least rows rows.
boolean	equals(Object obj)	Checks if the given object is equal to this dataset.
boolean	existsMissingValue()	existsMissingValue.
int	getColumn(Node variable)	getColumn.
Map<String,String>	getColumnToTooltip()	Getter for the field columnToTooltip.
Matrix	getCorrelationMatrix()	getCorrelationMatrix.
Matrix	getCovarianceMatrix()	getCovarianceMatrix.
DataBox	getDataBox()	Getter for the field dataBox.
double	getDouble(int row, int column)	getDouble.
Matrix	getDoubleData()	getDoubleData.
int	getInt(int row, int column)	getInt.
Knowledge	getKnowledge()	Getter for the field knowledge.
String	getName()	Gets the name of the data set.
NumberFormat	getNumberFormat()	getNumberFormat.
int	getNumColumns()	getNumColumns.
int	getNumRows()	getNumRows.
Object	getObject(int row, int col)	getObject.
int[]	getSelectedIndices()	getSelectedIndices.
Node	getVariable(int col)	getVariable.
Node	getVariable(String varName)	getVariable.
List<String>	getVariableNames()	getVariableNames.
List<Node>	getVariables()	Getter for the field variables.
boolean	isContinuous()	isContinuous.
boolean	isDiscrete()	isDiscrete.
boolean	isMixed()	isMixed.
boolean	isSelected(Node variable)	isSelected.
DataSet	like()	Returns a dataset with the same dimensions as this dataset, but with no data.
void	permuteRows()	Randomly permutes the rows of the dataset.
void	removeCols(int[] cols)	Removes the given columns from the data set.
void	removeColumn(int index)	Removes the variable (and data) at the given index.
void	removeColumn(Node variable)	Removes the given variable, along with all of its data.
void	removeRows(int[] selectedRows)	Removes the specified rows from the dataBox, updates the selection, multipliers, and knowledge accordingly.
static BoxDataSet	serializableInstance()	Generates a simple exemplar of this class to test serialization.
void	setDouble(int row, int column, double value)	Sets the value at the given (row, column) to the given double value, assuming the variable for the column is continuous.
void	setInt(int row, int column, int value)	Sets the value at the given (row, column) to the given int value, assuming the variable for the column is discrete.
void	setKnowledge(Knowledge knowledge)	Sets knowledge to a copy of the given object.
void	setName(String name)	Sets the name of the data model (may be null).
void	setNumberFormat(NumberFormat nf)	The number formatter used to print out continuous values.
void	setObject(int row, int col, Object value)	Sets the value at the given (row, column) to the given value.
void	setOutputDelimiter(Character character)	The character used a delimiter when the dataset is output
void	setSelected(Node variable, boolean selected)	Marks the given column as selected if 'selected' is true or deselected if 'selected' is false.
DataSet	subsetColumns(int[] indices)	Creates a new DataSet object containing only the specified columns.
DataSet	subsetColumns(List<Node> vars)	Creates and returns a dataset consisting of those variables in the list vars.
DataSet	subsetRows(int[] rows)	Creates a subset of rows from the existing DataSet.
DataSet	subsetRowsColumns(int[] rows, int[] columns)	Generates a subset of the current DataSet by selecting specified rows and columns.
String	toString()	toString.

Methods inherited from class java.lang.Object

getClass, hashCode, notify, notifyAll, wait, wait, wait

Constructor Details

BoxDataSet

public BoxDataSet(DataBox dataBox,
 List<Node> variables)

Constructs a new data set with the given number of rows and columns, with all values set to missing.

Parameters:

dataBox - The data box.

variables - The variables.

BoxDataSet

public BoxDataSet(BoxDataSet dataSet)

Makes of copy of the given data set.

Parameters:

dataSet - The data set to copy.

Method Details

serializableInstance

public static BoxDataSet serializableInstance()

Generates a simple exemplar of this class to test serialization.

Returns:

A simple exemplar of this class.

getColumnToTooltip

public Map<String,String> getColumnToTooltip()

Getter for the field columnToTooltip.

Specified by:

getColumnToTooltip in interface DataSet

Returns:

a Map object

getName

public String getName()

Gets the name of the data set.

Specified by:

getName in interface DataModel

Specified by:

getName in interface DataSet

Returns:

a String object

setName

public void setName(String name)

Sets the name of the data model (may be null).

Sets the name of the data set.

Specified by:

setName in interface DataModel

Parameters:

name - the name to set

getNumColumns

public int getNumColumns()

getNumColumns.

Specified by:

getNumColumns in interface DataSet

Returns:

the number of variables in the data set.

getNumRows

public int getNumRows()

getNumRows.

Specified by:

getNumRows in interface DataSet

Returns:

the number of rows in the rectangular data set, which is the maximum of the number of rows in the list of wrapped columns.

setInt

public void setInt(int row,
 int column,
 int value)

Sets the value at the given (row, column) to the given int value, assuming the variable for the column is discrete.

Sets the value at the given (row, column) to the given int value, assuming the variable for the column is discrete.

Specified by:

setInt in interface DataSet

Parameters:

row - The index of the case.

column - The index of the variable.

value - The value to set.

setDouble

public void setDouble(int row,
 int column,
 double value)

Sets the value at the given (row, column) to the given double value, assuming the variable for the column is continuous.

Sets the value at the given (row, column) to the given double value, assuming the variable for the column is continuous.

Specified by:

setDouble in interface DataSet

Parameters:

row - The index of the case.

column - The index of the variable.

value - The value to set.

getObject

public Object getObject(int row,
 int col)

getObject.

Specified by:

getObject in interface DataSet

Parameters:

row - The index of the case.

col - The index of the variable.

Returns:

the value at the given row and column as an Object. The type returned is deliberately vague, allowing for variables of any type. Primitives will be returned as corresponding wrapping objects (for example, doubles as Doubles).

setObject

public void setObject(int row,
 int col,
 Object value)

Sets the value at the given (row, column) to the given value.

Specified by:

setObject in interface DataSet

Parameters:

row - The index of the case.

col - The index of the variable.

value - The value to set.

getSelectedIndices

public int[] getSelectedIndices()

getSelectedIndices.

Specified by:

getSelectedIndices in interface DataSet

Returns:

the indices of the currently selected variables.

addVariable

public void addVariable(Node variable)

Adds the given variable to the data set.

Adds the given variable to the data set, increasing the number of columns by one, moving columns i >= index to column i + 1, and inserting a column of missing values at column i.

Specified by:

addVariable in interface DataSet

Parameters:

variable - The variable to add.

addVariable

public void addVariable(int index,
 Node variable)

Adds the given variable at the given index.

Adds the given variable to the dataset, increasing the number of columns by one, moving columns i >= index to column i + 1, and inserting a column of missing values at column i.

Specified by:

addVariable in interface DataSet

Parameters:

index - The index at which to add the variable.

variable - The variable to add.

getVariable

public Node getVariable(int col)

getVariable.

getVariable.

Specified by:

getVariable in interface DataSet

Parameters:

col - The index of the variable.

Returns:

the variable at the given column.

getColumn

public int getColumn(Node variable)

getColumn.

Specified by:

getColumn in interface DataSet

Parameters:

variable - The variable to check.

Returns:

the column index of the given variable.

changeVariable

public void changeVariable(Node from,
 Node to)

Changes the variable for the given column from from to to. Supported currently only for discrete variables.

Changes the variable for the given column from from to to. Supported currently only for discrete variables.

Specified by:

changeVariable in interface DataSet

Parameters:

from - The variable to change.

to - The variable to change to.

getVariable

public Node getVariable(String varName)

getVariable.

Specified by:

getVariable in interface DataModel

Specified by:

getVariable in interface DataSet

Parameters:

varName - a String object

Returns:

the variable with the given name, or null if no such variable exists.

getVariables

public List<Node> getVariables()

Getter for the field variables.

Specified by:

getVariables in interface DataSet

Specified by:

getVariables in interface VariableSource

Returns:

(a copy of) the List of Variables for the data set, in the order of their columns.

getKnowledge

public Knowledge getKnowledge()

Getter for the field knowledge.

Specified by:

getKnowledge in interface KnowledgeTransferable

Returns:

a copy of the knowledge associated with this data set. (Cannot be null.)

setKnowledge

public void setKnowledge(Knowledge knowledge)

Sets knowledge to a copy of the given object.

Sets knowledge to be associated with this data set. May not be null.

Specified by:

setKnowledge in interface KnowledgeTransferable

Parameters:

knowledge - the knowledge to set

getVariableNames

public List<String> getVariableNames()

getVariableNames.

Specified by:

getVariableNames in interface DataSet

Specified by:

getVariableNames in interface VariableSource

Returns:

(a copy of) the List of Variables for the data set, in the order of their columns.

setSelected

public void setSelected(Node variable,
 boolean selected)

Marks the given column as selected if 'selected' is true or deselected if 'selected' is false.

Marks the given column as selected if 'selected' is true or deselected if 'selected' is false.

Specified by:

setSelected in interface DataSet

Parameters:

variable - The variable to select or deselect.

selected - True to select the variable, false to deselect it.

clearSelection

public void clearSelection()

Marks all variables as deselected.

Specified by:

clearSelection in interface DataSet

ensureRows

public void ensureRows(int rows)

Ensures that the dataset has at least rows rows. Used for pasting data into the dataset.

Ensures that the dataset has at least the number of rows, adding rows if necessary to make that the case. The new rows will be filled with missing values.

Specified by:

ensureRows in interface DataSet

Parameters:

rows - The number of rows to ensure.

ensureColumns

public void ensureColumns(int columns,
 List<String> excludedVariableNames)

Ensures that the dataset has at least columns columns. Used for pasting data into the dataset. When creating new columns, names in the excludedVariableNames list may not be used. The purpose of this is to allow these names to be set later by the calling class, without incurring conflicts.

Ensures that the dataset has at least the given number of columns, adding continuous variables with unique names until that is true. The new columns will be filled with missing values.

Specified by:

ensureColumns in interface DataSet

Parameters:

columns - The number of columns to ensure.

excludedVariableNames - The names of variables that should not be used for new columns.

existsMissingValue

public boolean existsMissingValue()

existsMissingValue.

Specified by:

existsMissingValue in interface DataSet

Returns:

true if and only if this data set contains at least one missing value.

isSelected

public boolean isSelected(Node variable)

isSelected.

Specified by:

isSelected in interface DataSet

Parameters:

variable - The variable to check.

Returns:

true iff the given column has been marked as selected.

removeColumn

public void removeColumn(int index)

Removes the variable (and data) at the given index.

Removes the column for the variable at the given index, reducing the number of columns by one.

Specified by:

removeColumn in interface DataSet

Parameters:

index - The index of the variable to remove.

removeColumn

public void removeColumn(Node variable)

Removes the given variable, along with all of its data.

Removes the columns for the given variable from the dataset, reducing the number of columns by one.

Specified by:

removeColumn in interface DataSet

Parameters:

variable - The variable to remove.

subsetColumns

public DataSet subsetColumns(List<Node> vars)

Creates and returns a dataset consisting of those variables in the list vars. Vars must be a subset of the variables of this DataSet. The ordering of the elements of vars will be the same as in the list of variables in this DataSet.

Creates and returns a dataset consisting of those variables in the list vars. Vars must be a subset of the variables of this DataSet. The ordering of the elements of vars will be the same as in the list of variables in this DataSet.

Specified by:

subsetColumns in interface DataSet

Parameters:

vars - The variables to include in the new data set.

Returns:

a new data set consisting of the variables in the list vars.

isContinuous

public boolean isContinuous()

isContinuous.

Specified by:

isContinuous in interface DataModel

Specified by:

isContinuous in interface DataSet

Returns:

true iff this is a continuous data set--that is, if every column in it is continuous. (By implication, empty datasets are both discrete and continuous.)

isDiscrete

public boolean isDiscrete()

isDiscrete.

Specified by:

isDiscrete in interface DataModel

Specified by:

isDiscrete in interface DataSet

Returns:

true iff this is a discrete data set--that is, if every column in it is discrete. (By implication, empty datasets are both discrete and continuous.)

isMixed

public boolean isMixed()

isMixed.

Specified by:

isMixed in interface DataModel

Specified by:

isMixed in interface DataSet

Returns:

true if this is a mixed data set--that is, if it contains at least one continuous column and one discrete columnn.

getCorrelationMatrix

public Matrix getCorrelationMatrix()

getCorrelationMatrix.

Specified by:

getCorrelationMatrix in interface DataSet

Returns:

the correlation matrix for this dataset. Defers to Statistic.covariance() in the COLT matrix library, so it inherits the handling of missing values from that library--that is, any off-diagonal correlation involving a column with a missing value is Double.NaN, although all of the on-diagonal elements are 1.0. If that's not the desired behavior, missing values can be removed or imputed first.

getCovarianceMatrix

public Matrix getCovarianceMatrix()

getCovarianceMatrix.

Specified by:

getCovarianceMatrix in interface DataSet

Returns:

the covariance matrix for this dataset. Defers to Statistic.covariance() in the COLT matrix library, so it inherits the handling of missing values from that library--that is, any covariance involving a column with a missing value is Double.NaN. If that's not the desired behavior, missing values can be removed or imputed first.

getInt

public int getInt(int row,
 int column)

getInt.

Specified by:

getInt in interface DataSet

Parameters:

row - The index of the case.

column - The index of the variable.

Returns:

the value at the given row and column as an int, rounding if necessary. For discrete variables, this returns the category index of the datum for the variable at that column. Returns DiscreteVariable.MISSING_VALUE for missing values.

getDouble

public double getDouble(int row,
 int column)

getDouble.

Specified by:

getDouble in interface DataSet

Parameters:

row - The index of the case.

column - The index of the variable.

Returns:

the value at the given row and column as a double. For discrete data, returns the integer value cast to a double.

toString

public String toString()

toString.

Specified by:

toString in interface DataModel

Specified by:

toString in interface DataSet

Overrides:

toString in class Object

Returns:

a string, suitable for printing, of the dataset. Lines are separated by '\n', tokens in the line by whatever character is set in the setOutputDelimiter() method. The list of variables is printed first, followed by one line for each case. This method should probably not be used for saving to files. If that's your goal, use the DataSavers class instead.

See Also:

• setOutputDelimiter(Character)
• DataWriter

getDoubleData

public Matrix getDoubleData()

getDoubleData.

Specified by:

getDoubleData in interface DataSet

Returns:

a copy of the underlying COLT TetradMatrix matrix, containing all of the data in this dataset, discrete data included. Discrete data will be represented by ints cast to doubles. Rows in this matrix are cases, and columns are variables. The list of variable, in the order in which they occur in the matrix, is given by getVariables().

If isMultipliersCollapsed() returns false, multipliers in the dataset are first expanded before returning the matrix, so the number of rows in the returned matrix may not be the same as the number of rows in this dataset.

Throws:

IllegalStateException - if this is not a continuous data set.

See Also:

• getVariables()

subsetColumns

public DataSet subsetColumns(int[] indices)

Creates a new DataSet object containing only the specified columns.

Specified by:

subsetColumns in interface DataSet

Parameters:

indices - An array of integers representing the indices of the columns to include in the subset.

Returns:

A new DataSet containing only the specified columns.

subsetRows

public DataSet subsetRows(int[] rows)

Creates a subset of rows from the existing DataSet.

Specified by:

subsetRows in interface DataSet

Parameters:

rows - An array of integers representing the indices of the rows to be included in the subset.

Returns:

A new DataSet object containing only the specified rows from the original DataSet.

subsetRowsColumns

public DataSet subsetRowsColumns(int[] rows,
 int[] columns)

Generates a subset of the current DataSet by selecting specified rows and columns.

Specified by:

subsetRowsColumns in interface DataSet

Parameters:

rows - an array of row indices to include in the subset

columns - an array of column indices to include in the subset

Returns:

a new DataSet object containing only the specified rows and columns

removeCols

public void removeCols(int[] cols)

Removes the given columns from the data set.

Removes the given columns from the data set.

Specified by:

removeCols in interface DataSet

Parameters:

cols - The indices of the columns to remove.

removeRows

public void removeRows(int[] selectedRows)

Removes the specified rows from the dataBox, updates the selection, multipliers, and knowledge accordingly.

Specified by:

removeRows in interface DataSet

Parameters:

selectedRows - an array of integers representing the indices of the rows to be removed from the dataBox

equals

public boolean equals(Object obj)

Checks if the given object is equal to this dataset.

Specified by:

equals in interface DataSet

Overrides:

equals in class Object

Parameters:

obj - The object to check.

Returns:

True if the given object is equal to this dataset.

copy

public DataSet copy()

Returns a copy of this dataset.

Specified by:

copy in interface DataModel

Specified by:

copy in interface DataSet

Returns:

A copy of this dataset.

like

public DataSet like()

Returns a dataset with the same dimensions as this dataset, but with no data.

Specified by:

like in interface DataSet

Returns:

a dataset with the same dimensions as this dataset, but with no data.

setOutputDelimiter

public void setOutputDelimiter(Character character)

The character used a delimiter when the dataset is output

Sets the character ('\t', ' ', ',', for instance) that is used to delimit tokens when the data set is printed out using the toString() method.

Specified by:

setOutputDelimiter in interface DataSet

Parameters:

character - The character used as a delimiter when the dataset is output

See Also:

• toString()

permuteRows

public void permuteRows()

Randomly permutes the rows of the dataset.

Specified by:

permuteRows in interface DataSet

getNumberFormat

public NumberFormat getNumberFormat()

getNumberFormat.

Specified by:

getNumberFormat in interface DataSet

Returns:

the number format, which by default is the one at NumberFormatUtil.getInstance().getNumberFormat(), but can be set by the user if desired.

See Also:

• setNumberFormat(java.text.NumberFormat)

setNumberFormat

public void setNumberFormat(NumberFormat nf)

The number formatter used to print out continuous values.

Sets the number format to be used when printing out the data set. The default is the one at

Specified by:

setNumberFormat in interface DataSet

Parameters:

nf - The number formatter used to print out continuous values.

getDataBox

public DataBox getDataBox()

Getter for the field dataBox.

Returns:

the data box that holds the data for this data set.