mml

Class Tree.Est

java.lang.Object

la.la.Value

la.la.Function

la.la.Function.Native

mml.Estimator

mml.UPModel.Est

mml.UPFunctionModel.Est

mml.Tree.Est

All Implemented Interfaces:

java.lang.Comparable<Value>

Enclosing class:

Tree

public class Tree.Est
extends UPFunctionModel.Est

The class of Estimators for a Tree FunctionModel. It is public so that it can be extended, e.g., to allow for new kinds of Tree.Fork or changes to the search, etc..

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
class	Tree.Est.Sel Class of variable selector of the input datum id, where iod=(id,od).

Nested classes/interfaces inherited from class la.la.Function.Native

Function.Native.WithInverse

Nested classes/interfaces inherited from class la.la.Function

Function.Cts2Cts, Function.Cts2Cts2Cts, Function.CtsD2CtsD, Function.HasInverse, Function.Native, Function.Native2, Function.Native3

Nested classes/interfaces inherited from class la.la.Value

Value.Atomic, Value.Bool, Value.Char, Value.Chars, Value.Cts, Value.Defer, Value.Discrete, Value.Enum, Value.Inc_Or, Value.Int, Value.Lambda, Value.List, Value.Maybe, Value.Option, Value.Real, Value.Scannable, Value.Structured, Value.Triv, Value.Tuple

Field Summary

Fields

Modifier and Type	Field and Description
Estimator	leafEst leafEst estimates a Model of the output datum, col=1.
Tree.Est.Sel	sel0 Selects all of the input datum id, where iod=(id,od).

Fields inherited from class mml.Estimator

params

Fields inherited from class la.la.Value

C, comparator, CR, E, eight, ffalse, five, fiveR, four, fourR, half, negCR, negOne, negOneR, negTenR, nilVal, nine, None, one, oneR, PI, PIby2, point01, point1, point5, point9, quarter, seven, six, ten, tenR, three, threeR, triv, ttrue, two, twoR, zero, zeroR

Constructor Summary

Constructors

Constructor and Description
Est(Value ps) Note, ps is passed to leafUPM.estimator(ps) to make leafEst.

Method Summary

Modifier and Type	Method and Description
Tree.DFork	estDF(int nF, boolean[] f, Vector col, int cN, Tree.Est.Sel sel, Vector ds, int lookahead) Estimate a DFork on the Discrete Bounded column 'col', numbered 'cN', of the input (indep.) datum.
Tree.Leaf	estLeaf(Vector ds) Estimate a Leaf, i.e., Model of the output (dependent) column ignoring the input (indep.) column(s).
Tree.M	estOF(int nF, boolean[] f, Vector col, int cN, Tree.Est.Sel sel, Vector ds, int lookahead) Estimate an OFork on the input column 'col', numbered 'cN'.
Tree.M	recSearch(int nF, boolean[] f, Vector ds, int lookahead) Coordinate the recursive search for a good (best?) Tree Model.
Tree.M	search(Vector ds, int lookahead) Given a data-set, ds=(id,od)*, estimate a fully parameterised M:id→od; calls recSearch.
Tree.M	searchMV(int nF, boolean[] f, Vector ds, int lookahead, Tree.M leaf) Search for a column of a Multivariate input datum on which to split.
Tree.Est.Sel	sel(int col) Selector of input column (variable) 'col' of the input datum id, where iod=(id,od).
Tree.M	ss2Model(Value ss) Given statistics ss (= ds) of a data-set, ds, estimate a fully parameterised M:id→od; calls search(ss, 1).

Methods inherited from class mml.UPFunctionModel.Est

ds2FunctionModel, ds2Model, ss2FunctionModel

Methods inherited from class mml.UPModel.Est

defnParams, sp2Model, stats, stats, toString

Methods inherited from class mml.Estimator

apply, asUPModel, ds2Model, ds2ModelSp, ss2ModelSp, stats, stats

Methods inherited from class la.la.Function

bOp, fromBop, fromUop, main, type, uOp

Methods inherited from class la.la.Value

AoM, bOp, compareTo, cons, cts, elt, errMsg, error, force, isTuple, just, n, nElts, nlAoM, pair, print, quad, real, RTE, triple, tuple, UOE, x

Methods inherited from class java.lang.Object

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

Field Detail

leafEst

public final Estimator leafEst

leafEst estimates a Model of the output datum, col=1. Called by estLeaf.

sel0

public Tree.Est.Sel sel0

Selects all of the input datum id, where iod=(id,od).

Constructor Detail

Est

public Est(Value ps)

Note, ps is passed to leafUPM.estimator(ps) to make leafEst.

Method Detail

ss2Model

public Tree.M ss2Model(Value ss)

Given statistics ss (= ds) of a data-set, ds, estimate a fully parameterised M:id→od; calls search(ss, 1). You can call the latter directly to vary the lookahead.

Specified by:

ss2Model in class UPFunctionModel.Est

search

public Tree.M search(Vector ds,
                     int lookahead)

Given a data-set, ds=(id,od)^*, estimate a fully parameterised M:id→od; calls recSearch. If lookahead=0, potential Fork choices will be compared initially on the basis that each one is a Fork of Leafs. Increasing lookahead >0 may give a better result, but will take more time. Called by ss2Model(ss).

recSearch

public Tree.M recSearch(int nF,
                        boolean[] f,
                        Vector ds,
                        int lookahead)

Coordinate the recursive search for a good (best?) Tree Model. There remain 'nF' free Fork-able input columns (as indicated in f, if f≠null). Calls estDF, estOF, and searchMV. Called by search.

searchMV

public Tree.M searchMV(int nF,
                       boolean[] f,
                       Vector ds,
                       int lookahead,
                       Tree.M leaf)

Search for a column of a Multivariate input datum on which to split. If there are multiple candidates, try each one in turn, using lookahead-1, to find the likely best one. Calls sMV_cN. Called by searchMV.

sel

public Tree.Est.Sel sel(int col)

Selector of input column (variable) 'col' of the input datum id, where iod=(id,od).

estDF

public Tree.DFork estDF(int nF,
                        boolean[] f,
                        Vector col,
                        int cN,
                        Tree.Est.Sel sel,
                        Vector ds,
                        int lookahead)

Estimate a DFork on the Discrete Bounded column 'col', numbered 'cN', of the input (indep.) datum. Calls recSearch.

estOF

public Tree.M estOF(int nF,
                    boolean[] f,
                    Vector col,
                    int cN,
                    Tree.Est.Sel sel,
                    Vector ds,
                    int lookahead)

Estimate an OFork on the input column 'col', numbered 'cN'. Specifying the median as the splitting Value would cost 1-bit, a quartile 3-bits, an octile 5 bits, and so on, thus effecting variable precision (Wallace & Patrick 1993). Calls recSearch.

estLeaf

public Tree.Leaf estLeaf(Vector ds)

Estimate a Leaf, i.e., Model of the output (dependent) column ignoring the input (indep.) column(s). Note, for a k-way tree #leaves = (k-1) × #forks + 1.