Package edu.cmu.tetrad.bayes

Class CptMapCounts

java.lang.Object

edu.cmu.tetrad.bayes.CptMapCounts

All Implemented Interfaces:

CptMap, TetradSerializable, Serializable

public class CptMapCounts
extends Object
implements CptMap

Represents a conditional probability table (CPT) in a Bayes net. This represents the CPT as a map from a unique integer index for a particular node to the cell count for that node, where 0's are not stored. Row counts are also stored, so that the probability of a cell can be calculated. A prior cell count of 0 is assumed for all cells, but this may be set by the user to any non-negative count. (A prior count of 0 is equivalent to a maximum likelihood estimate.)

Author:

josephramsey

See Also:

• Serialized Form

Constructor Summary

Constructors

Constructor	Description
CptMapCounts(int numRows, int numColumns)	Constructs a new probability map, a map from a unique integer index for a particular node to the probability of that node taking on that value, where NaN's are not stored.
CptMapCounts(DataSet data)	Constructs a new CptMap based on counts from a given dataset.

Method Summary

Modifier and Type	Method	Description
void	addCounts(int row, int column, int count)	Adds the specified count to the cell count at the given row and column.
double	get(int row, int column)	Returns the probability of the node taking on the value specified by the given row and column.
int	getNumColumns()	Returns the number of columns in the probability map.
int	getNumRows()	Returns the number of rows in the probability map.
double	getPriorCount()	Retrieves the prior count for all cells in the CptMapCounts.
void	setPriorCount(double priorCount)	Sets the prior count for all cells in the CptMapCounts.

Methods inherited from class java.lang.Object

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

Constructor Details

CptMapCounts

public CptMapCounts(int numRows,
 int numColumns)

Constructs a new probability map, a map from a unique integer index for a particular node to the probability of that node taking on that value, where NaN's are not stored. This probability map assumes that there is a certain number of rows and a certain number of columns in the table.

Parameters:

numRows - the number of rows in the table

numColumns - the number of columns in the table

CptMapCounts

public CptMapCounts(DataSet data)

Constructs a new CptMap based on counts from a given dataset.

Parameters:

data - the DataSet object representing the probability matrix

Throws:

IllegalArgumentException - if the data set is null or not discrete

Method Details

get

public double get(int row,
 int column)

Returns the probability of the node taking on the value specified by the given row and column.

Specified by:

get in interface CptMap

Parameters:

row - the row of the node

column - the column of the node

Returns:

the probability of the node taking on the value specified by the given row and column

addCounts

public void addCounts(int row,
 int column,
 int count)

Adds the specified count to the cell count at the given row and column.

Parameters:

row - the row index of the cell count

column - the column index of the cell count

count - the count to be added to the cell count

Throws:

IllegalArgumentException - if the row or column is out of bounds

getNumRows

public int getNumRows()

Returns the number of rows in the probability map.

Specified by:

getNumRows in interface CptMap

Returns:

the number of rows in the probability map.

getNumColumns

public int getNumColumns()

Returns the number of columns in the probability map.

Specified by:

getNumColumns in interface CptMap

Returns:

the number of columns in the probability map.

getPriorCount

public double getPriorCount()

Retrieves the prior count for all cells in the CptMapCounts.

Returns:

the prior count for all cells in the CptMapCounts.

setPriorCount

public void setPriorCount(double priorCount)

Sets the prior count for all cells in the CptMapCounts. The prior count is used in parameter estimation for Bayesian networks.

Parameters:

priorCount - the value to set as the prior count.