Processing math: 100%

Parameters.ox

Parameter is the base class for values to optimize over.

Parameter is derived from Quantity to represent variables that are continuous. (Variables in DDP such as states and actions are Discrete.)

THe value of a Parameter is the v member of the class inherited from Objective. Other members are used to store starting values and to restrict the range the value takes on during optimization.

Click on Hierarchy to see the different kinds of parameters in a logical order.

ParameterBLocks are a derived class for reprsenting a vector of parameters treated together.

Parameters to maxmiize over are added to the Objective before applying an Algorithm which modifies the parameter values.

Global functions

Functions
NormalDistParmeters		Create an array of Normal Distribution parameters and return it.
TransitionMatrix		Create an array of Simplex blocks that act as a transition matrix.

Bounded : Limited : Parameter : Quantity

A parameter contained in an open interval

$L \lt v \lt U$ .

Scale, free value and transformation

$$\eqalign{ s &= \log\left( {v_0-L \over U-v_0} \right)\cr f &= 1\cr v &= L+UB {\exp\{sf\}\over 1+\exp\{sf\}\cr }$$

Warning issued if |v₀)-(UB+UB)/2| < NearFlat

Public fields
LB const AV compatible, the lower bound

UB const AV compatible, the upper bound

Public methods
Bounded Create a parameter bounded above and below.

Inherited methods from Limited:: Limited
Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from Limited:: nearflat
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

BoundedAbove : Limited : Parameter : Quantity

Parameter bounded from above

$-\infty \lt v \lt U$ .

Scale, free value and transformation

$\eqalign{ s &= \log(U-v_0)\cr f &= 1\cr v &= U+\exp\{sf\}\cr }$

Warning issued if |v₀)-UB + 1| < NearFlat

Public fields
UB const AV compatible, the upper bound

Public methods
BoundedAbove Create a new parameter bounded above.

Menu virtual

Inherited methods from Limited:: Limited
Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from Limited:: nearflat
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

BoundedBelow : Limited : Parameter : Quantity

A parameter bounded from below:

$L \lt v \lt \infty$ .

Scale, free value and transformation

$\eqalign{ s &= \log(v_0-L)\cr f &= 1\cr v &= L+\exp\{sf\}\cr }$

Warning issued if |v₀)-LB-1| < NearFlat

Public fields
LB const AV compatible, the lower bound

Public methods
BoundedBelow A parameter bounded below.

Inherited methods from Limited:: Limited
Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from Limited:: nearflat
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Coefficients : ParameterBlock : Parameter : Quantity

Vector of free parameters.

x^j = Free()

Public methods
Coefficients		Create a block of free parameters.

Inherited methods from ParameterBlock:: AddToBlock, BlockCode, Encode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Correlation : Bounded : Limited : Parameter : Quantity

Bounded as: -1 < v < 1 .

Equivalence

Correlation(L,v0) ≡ Bounded(L,-1.0,1.0,v0))

Public methods
Correlation		Create a new parameter bounded above by 1 and below by -1.

Inherited methods from Bounded:: Bounded
Inherited methods from Limited:: Limited
Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from Bounded:: LB, UB
Inherited fields from Limited:: nearflat
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Decreasing : Ordered : ParameterBlock : Parameter : Quantity

Vector of parameters that are sequentially decreasing.

x⁰ = BoundedAbove(UB) or Free()
x^j = BoundedAbove(x^j-1)

Public methods
Decreasing		Create a decreasing vector of parameters.

Inherited methods from Ordered:: Ordered
Inherited methods from ParameterBlock:: AddToBlock, BlockCode, Encode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from Ordered:: Anchored, B
Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

DecreasingReturns : ParameterBlock : Parameter : Quantity

Vector of J probabilities that sum to strictly less than 1.

Relationship

x⁰ = Probability
...
x^j = Bounded(0.0,1-∑_k=0…j-1 x^k)

Public methods
DecreasingReturns		Create a block of decreasing returns to scale coefficients.

Inherited methods from ParameterBlock:: AddToBlock, Encode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Determined : Parameter : Quantity

Value determined exactly by some other value, not chosen by optimization.

Public methods
Determined		Create a fixed, pre-determined parameter.
SetDoNotVary		Do nothing because DoNotVary does not change for Determined parameter.
ToggleDoNotVary		Do nothing because DoNotVary does not toggle for Determined parameter.

Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

FixedBlock : ParameterBlock : Parameter : Quantity

Vector of values determined exactly by some other value.

Public methods
Decode
Encode		Default encoding: no scaling, no constraining.
FixedBlock		create a new vector of parameters that are not changed by algorithms.

Inherited methods from ParameterBlock:: AddToBlock, BlockCode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Free : Parameter : Quantity

Can take on any real number:

$-\infty \lt v \lt \infty$ .

Scale, free value and transformation

$$s = \cases{ v_0 & $|v_0| \gt

$NearFlat\cr 1 & else\cr}$ \(

\)$f = \cases{ 1 & $|v_0| \gt $NearFlat\cr v_0 & else\cr}$ \(

\) $v = sf$ $

Public methods
Encode virtual .

Free Create a free, unrestricted parameter.

Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Increasing : Ordered : ParameterBlock : Parameter : Quantity

Vector of parameters that are sequentially increasing.

x⁰ = BoundedBelow(LB) or Free()
x^j = BoundedBelow(x^j-1)

Public methods
Increasing		Create an increasing vector of parameters.

Inherited methods from Ordered:: Ordered
Inherited methods from ParameterBlock:: AddToBlock, BlockCode, Encode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from Ordered:: Anchored, B
Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Limited : Parameter : Quantity

Container for parameters are neither Free nor Determined.

Public fields
nearflat		starting value is near the flat spot of the transformation.
Public methods
Limited		.

Inherited methods from Parameter:: Decode, Encode, Menu, Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Negative : BoundedAbove : Limited : Parameter : Quantity

Bounded from above by 0.

Equivalence

Negative(L,v0) ≡ BoundedAbove(L,0.0)

Public methods
Negative		Create a parameter bounded above by 0.

Inherited methods from BoundedAbove:: BoundedAbove, Menu
Inherited methods from Limited:: Limited
Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from BoundedAbove:: UB
Inherited fields from Limited:: nearflat
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Ordered : ParameterBlock : Parameter : Quantity

Public fields
Anchored	const	list is anchored .
B	const	AV compatible, bound for first value.
Public methods
Ordered		Create a block with parameters that maintain an order.

Inherited methods from ParameterBlock:: AddToBlock, BlockCode, Encode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

ParameterBlock : Parameter : Quantity

Two or more parameters whose ranges interact or are related for some other reason.

Public fields
curpar		current param in V.
N		length of block.
Psi		array of parameters .
PsiL		array of labels.
Public methods
AddToBlock		Append parameter to the block.
BlockCode	virtual	Default update for an unstructured block of parameters.
Encode	virtual	Default encoding: no scaling, no constraining.
Menu	virtual
ParameterBlock		Create a new block of related Parameters.
SetDoNotVary		Set the value of DoNotVary.
ToggleDoNotVary		Toggle DoNotVary for All or some block elements.
Xb		Return $X\beta = X*CV(beta)$ .

Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Positive : BoundedBelow : Limited : Parameter : Quantity

Parameter Bounded from below by 0.

Equivalence

Positive(L,v0) ≡ BoundedBelow(L,0,v0)

Public methods
Positive		Create a parameter bounded below by 0.

Inherited methods from BoundedBelow:: BoundedBelow
Inherited methods from Limited:: Limited
Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from BoundedBelow:: LB
Inherited fields from Limited:: nearflat
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Probabilities : ParameterBlock : Parameter : Quantity

Vector of Probabilities.

x^j = Probability()

Public methods
Probabilities		Create a block of parameters all constrained to be probabilities.

Inherited methods from ParameterBlock:: AddToBlock, BlockCode, Encode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Probability : Bounded : Limited : Parameter : Quantity

Bounded as: 0 < v < 1 .

Equivalence

Probability(L,v0) &equiv Bounded(L,0.0,1.0)

Public methods
Probability		Create a new parameter bounded above by 1 and below by 0.

Inherited methods from Bounded:: Bounded
Inherited methods from Limited:: Limited
Inherited methods from Parameter:: Parameter, ReadCGI, ReInitialize, Reset, SetDoNotVary, ToggleDoNotVary
Inherited methods from Quantity:: SetVolume

Inherited fields from Bounded:: LB, UB
Inherited fields from Limited:: nearflat
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Simplex : ParameterBlock : Parameter : Quantity

Vector of J probabilities that sum to 1.

x⁰ = Probability
x^j = Bounded(0.0,1-∑_k=0…j-1 x^k)
⋮
x^J = Determined(1-∑x^j)

See also:: TransitionMatrix

Public fields
fval		final value, used for cumulating.
Public methods
Last	static
Simplex		Create a simplex of parameters.

Inherited methods from ParameterBlock:: AddToBlock, Encode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

StDeviations : ParameterBlock : Parameter : Quantity

Positive Vector.

x^j = Positive()

Public methods
StDeviations		Create a block of positive parameters.

Inherited methods from ParameterBlock:: AddToBlock, BlockCode, Encode, Menu, ParameterBlock, SetDoNotVary, ToggleDoNotVary, Xb
Inherited methods from Parameter:: Decode, Parameter, ReadCGI, ReInitialize, Reset
Inherited methods from Quantity:: SetVolume

Inherited fields from ParameterBlock:: curpar, N, Psi, PsiL
Inherited fields from Parameter:: block, DoNotConstrain, DoNotVary, f, ival, NearFlat, scale, start
Inherited fields from Quantity:: L, logf, pos, track, v, Volume

Global

NormalDistParmeters

NormalDistParmeters ( L , ivals )

Create an array of Normal Distribution parameters and return it.

This does not create a ParameterBlock. It creates separate parameter objects and returns the array of them. It can be sent to Normal() related functions/objects that require a pars vector or array.

Parameters:

L	Label prefix for the parameters
ivals	initial values vector see NormalParams default is <0.0;1.0>, which means parameters start as standard normal if ivals has three elements the last is the correlation coefficient $\rho$

See also:

NormalParams

TransitionMatrix

TransitionMatrix ( L , imat )

Create an array of Simplex blocks that act as a transition matrix.

Parameters:

L label

inmat

a square matrix of initial values for the transitions.

Each column of inmat should be a proper transition (not each row).

Example:

Create a 3x3 transition matrix, where the current state is the column and the next state is the row. Initialize the transition but make the transition an array of Simplex blocs that can be controlled by an optimization routine.

  decl tmat =< 0.90 ~ 0.09 ~0.05;
               0.02 ~ 0.80 ~0.3;
               0.08 ~ 0.01 ~0.11>

   decl m = TransitionMatrix("p",tmat);
   Parameters(m);

   println("The current Markov transition matrix is ", CV(m));

see Markov State Variable

Bounded

Bounded :: Bounded ( L , LB , UB , v0 )

Create a parameter bounded above and below.

Parameters:

L	parameter label
LB	static function, double or Parameter, the lower bound
UB	static function, double or Parameter, the upper bound
v0	CV compatible default value, v₀

LB

const decl LB [public]

AV compatible, the lower bound

UB

const decl UB [public]

AV compatible, the upper bound

BoundedAbove

BoundedAbove :: BoundedAbove ( L , UB , v0 )

Create a new parameter bounded above.

Parameters:

L	parameter label
LB	double or Parameter, the upper bound
v0	CV compatible default value, v₀

Comments:

If UB is a parameter then the bounding is dynamic. The parameter will always be strictly less than UB as UB's value changes.
UB should be added to the parameter list first.

UB

const decl UB [public]

AV compatible, the upper bound

BoundedBelow

BoundedBelow :: BoundedBelow ( L , LB , v0 )

A parameter bounded below.

Parameters:

L	parameter label
LB	double or CV-compatible object, the lower bound
v0	CV compatible default value, v₀

Comments:

If LB is a parameter then the bounding is dynamic. The parameter will always be strictly greater than LB as LB value changes.
LB should be added to the parameter list first.

LB

const decl LB [public]

AV compatible, the lower bound

Coefficients

Coefficients :: Coefficients ( L , ivals , labels )

Create a block of free parameters.

Parameters:

L	label for the block (e.g. the equation)
ivals	0: set to length of labels and initialize values to 0 > 0: number of coefficients, initialize to 0
labels	0: use numbers for labels array of strings, labels for coefficients (e.g. variable names) vector: initial values

Correlation

Correlation :: Correlation ( L , v0 )

Create a new parameter bounded above by 1 and below by -1.

Parameters:

L	parameter label
v0	CV compatible default value, v₀

See also:

Bounded

Decreasing

Decreasing :: Decreasing ( L , UB , ivals , Anchored )

Create a decreasing vector of parameters. LB < x₁ > x₂ > … > x_N

Parameters:

L	label
UB	upper bound for first variable, a double, CV comptible. send .Inf to make the first parameter free
ivals	integer, dimension of sequence OR N×1 vector of initial values
Anchored	TRUE, UB should be the first element of the block (as a Determined parameter)

Comments:

if ivals is an integer N then the sequence is initialized as UB-1, UB-2, … UB-N

DecreasingReturns

DecreasingReturns :: DecreasingReturns ( L , ivals )

Create a block of decreasing returns to scale coefficients.

Parameters:

label

ivals

N×1 vector, initial values

Cross-parameter restrictions

0< x_i < 1

∑_i=9…N-1  x_i < 1.

All elements

Determined

Determined :: Determined ( L , v0 )

Create a fixed, pre-determined parameter.

Parameters:

L	string
v0	double, initial value, current value determined by something else CV compatible default value, v₀

SetDoNotVary

Determined :: SetDoNotVary ( setting )

Do nothing because DoNotVary does not change for Determined parameter.

ToggleDoNotVary

Determined :: ToggleDoNotVary ( )

Do nothing because DoNotVary does not toggle for Determined parameter.

FixedBlock

Decode

Encode

FixedBlock :: Encode ( )

Default encoding: no scaling, no constraining.

v = start;
scale = 1.0;
f = v;

Returns:: v

FixedBlock

FixedBlock :: FixedBlock ( L , v0 )

create a new vector of parameters that are not changed by algorithms.

Parameters:

L	label
v0	vector or array of values.

Free

Encode

virtual Free :: Encode ( )

Free

Free :: Free ( L , v0 )

Create a free, unrestricted parameter.

Parameters:

L	parameter label
v0	CV compatible default value

Increasing

Increasing :: Increasing ( L , LB , ivals , Anchored )

Create an increasing vector of parameters.

LB < x₁ < x₂ < … < x_N

Parameters:

L	label
LB	lower bound for first variable, a double, Parameter or static function send -.Inf to make the first parameter free
ivals	integer, dimension of sequence OR N×1 vector of initial values
Anchored	TRUE, LB should be the first element of the block (as a Determined parameter)

Comments:

if ivals is an integer N then the sequence is initialized as LB+1, LB+2, … LB+N.

Limited

Limited :: Limited ( L , v0 )

nearflat

decl nearflat [public]

starting value is near the flat spot of the transformation.

Negative

Negative :: Negative ( L , v0 )

Create a parameter bounded above by 0.

Parameters:

L	parameter label
v0	CV compatible default value, v₀

Comments:

Equivalent to BoundedAbove(L,0.0,v0)

See also:

BoundedBelow

Ordered

Anchored

const decl Anchored [public]

list is anchored .

B

const decl B [public]

AV compatible, bound for first value.

Ordered

Ordered :: Ordered ( L , B , ivals , sign , Anchored )

Create a block with parameters that maintain an order.

Parameters:

L	label
B	CV-compatible bound or Determined parameter (can be -∞ or +∞)
ivals	vector of initial values scalar number of ordered elements (not counting Anchor)
sign	ordering
Anchored	if TRUE the first element of the block is B

ParameterBlock

AddToBlock

ParameterBlock :: AddToBlock ( ... )

Append parameter to the block.

Parameters:

psi	Parameter to add.
...	more parameters

BlockCode

virtual ParameterBlock :: BlockCode ( )

Default update for an unstructured block of parameters. The default is to simply increment curpar (and reset 0 when it reaches N).

curpar

decl curpar [public]

current param in V.

Encode

virtual ParameterBlock :: Encode ( )

Default encoding: no scaling, no constraining.

v = start;
scale = 1.0;
f = v;

Returns:: v

N

decl N [public]

length of block.

ParameterBlock

ParameterBlock :: ParameterBlock ( L , ... )

Create a new block of related Parameters.

Parameters:

L	string Label for block
...	a list of parameters to add to the block if only one argument is supplied it should be an array of parameters.

Psi

decl Psi [public]

array of parameters .

PsiL

decl PsiL [public]

array of labels.

SetDoNotVary

ParameterBlock :: SetDoNotVary ( setting )

Set the value of DoNotVary.

Parameters:

setting

TRUE or FALSE

ToggleDoNotVary

ParameterBlock :: ToggleDoNotVary ( elements )

Toggle DoNotVary for All or some block elements.

Parameters:

elements

DoAll or a vector of parameter indices

Xb

ParameterBlock :: Xb ( X )

Return

$X\beta = X*CV(beta)$ .

Parameters:

X	row vector or matrix conforming to the coefficient vector.

Returns:

$X*CV(this)$ .

Positive

Positive :: Positive ( L , v0 )

Create a parameter bounded below by 0.

Parameters:

L	parameter label
v0	CV compatible default value, v0

Comments:

Equivalent to BoundedBelow(L,0.0,ival)

Probabilities

Probabilities :: Probabilities ( L , ivals , labels )

Create a block of parameters all constrained to be probabilities.

Parameters:

L	label for the block
ivals	0: set to length of labels and initialize values to 0 > 0: number of postivie parameters, initialize to 1.0 vector: initial values
labels	0: use numbers for labels array of strings, labels for values (e.g. variable names)

See also:

Simplex

Probability

Probability :: Probability ( L , v0 )

Create a new parameter bounded above by 1 and below by 0.

Parameters:

L	parameter label
v0	CV compatible default value, v₀

See also:

Bounded

Simplex

fval

decl fval [public]

final value, used for cumulating.

Last

static Simplex :: Last ( )

Simplex

Simplex :: Simplex ( L , ivals )

Create a simplex of parameters.

Parameters:

label

ivals

integer: dimension of simplex initialized as equal values 1/N
N×1 vector, initial values

Cross-parameter restrictions

0 < x₀ < 1
0 < x₁ < x₀
⋮
0 < x_i < ∑_j=0…i-1  x_j
x_N-1 = 1 - ∑_j=0…N-2  x_j

The first N-1 elements of the block are Bounded parameters. The final one is Determined.

StDeviations

StDeviations :: StDeviations ( L , ivals , labels )

Create a block of positive parameters.

Parameters:

L	label for the block
ivals	0: set to length of labels and initialize all values at 1 > 0: number of positive parameters, initialize to 1.25 vector: initial values
labels	0: use numbers for labels array of strings, labels for values (e.g. variable names) For example, a vector of standard deviations.