since version 1.0
since version 1.0
since version 1.0
since version 1.0
since version 1.0
since version 1.0
since version 1.0
since version 1.0 - Does not use matrices. So, e.g. the expression 2x + 3y + x^2 + 2xy + 3y^2 has three quadratic coefficients: 1, 2, 3
since version 1.0
since version 2.0
We use the LPSOLVE definition of SOS, which goes back to a paper by J.A.Tomlin in "Studies on Graphs and Discrete Programming", 1981.
since version 2.0
experiment
since version 2.0
experiment
since version 2.0
since version 2.0
experiment
proposal
proposal
proposal
proposal
proposal
draft
experiment
experiment
experiment
finalization
finalization
realTimeData and simulations. experiment
experiment
finalization
since version 1.0
Number of the var child elements must agree with numberOfVariables (after "mult" has been considered)
semi-continuous (D) and semi-integer (J) types added in version 2.0
since version 1.0. Variable type "D" (semi-continuous) and "J" (semi-integer) since 2.0.
C - continuous
B- binary
I - integer
S - string
D - semi-continuous
J - semi-integer
since version 1.0
Number of the obj child elements must agree with numberOfObjectives (after "mult" has been considered)
since version 1.0
number of coef child elements must match numberOfObjCoef
If idx is repeated, the objective terms/coefficients are added together.
since version 1.0
since version 1.0
Number of the con child elements must agree with numberOfConstraints (after "mult" has been considered)
since version 1.0
since version 1.0
Number of entries in the start element must equal numberOfVariables + 1 (if columnMajor) or numberOfConstraints + 1 (if rowMajor), after considering "mult". Number of entries in the value and rowIdx/colIdx child elements must agree with numberOfValues (after "mult" has been considered). Multiple coefficients in the same location are added together.
since version 1.0
Number of the qTerm child elements must agree with numberOfQuadraticTerms
since version 1.0
index of the constraint or objective
index of the first variable
index of the second variable
coefficient of the
quadratic term
since version 1.0
Number of the child nl elements must agree with numberOfNonlinearExpressions
shape attribute added in version 2.0
since version 1.0
index of constraint or objective row
proposal
For complex-valued expressions only - real-valued expressions should be set up as nonlinearExpressions
draft
number must match numberOfComplexVar after "mult" has been considered
draft
must be constant matrix; zero if missing
contains var - lb; assumed all of C if missing
must be constant matrix; zero if missing
contains ub - var; assumed all of C if missing
same type for every newly created variable
draft
number must match numberOfComplexObj after "mult" has been considered
draft
e.g., if symmetry = "upper", only objectives in the upper triangle (including diagonal) are generated and passed to the solver interface.
matrix objective has a nonzero entry wherever the template matrix does also. If templateMatrixIdx missing: Completely dense matrix.
controls reuse of core objectives instead of creation of new ones.
Note: only where template matrix indicates a nonzero.
If missing, all objectives indicated by the use of the template matrix are created fresh.
x is (weakly) preferred to y if obj(x) - obj(y) lies in cone. If reference uses only core objectives, orderConeIdx can be omitted and defaults to the corresponding orthant cone
added to objective; zero if missing
draft
number must match numberOfComplexCon after "mult" has been considered
draft
e.g., if symmetry = "upper", only variables in the upper triangle (including diagonal) are generated and passed to the solver interface.
matrix constraint has a nonzero entry wherever the template matrix does also. If templateMatrixIdx missing: Completely dense matrix.
controls reuse of core constraints instead of creation of new ones.
Note: only where template matrix indicates a nonzero.
If missing, all constraints indicated by the use of the template matrix are created fresh.
must be constant matrix; zero if missing
contains mtx expr - lbMtx; assumed all R^n if missing
must be constant matrix; zero if missing
contains ubMtx - mtx expr; assumed all R^n if missing
draft
Number must agree with numberOfExpr
since version 1.0
refers to existing complex objectives (if negative) or complex constraints (if nonnegative)
draft
Number must agree with numberOfExpr
draft
refers to existing complex-valued objectives (if negative) or complex-valued constraints (if nonnegative)
since version 2.0
For matrix-valued expressions only - scalar-valued expressions should be set up as nonlinearExpressions
number must match numberOfMatrixVar after "mult" has been considered
number must agree with numberOfDescriptiors
can be used, e.g., to set up Toeplitz or Cauchy matrices
e.g., if symmetry = "upper", only variables in the upper triangle (including diagonal) are generated and passed to the solver interface.
matrix variable has a nonzero entry wherever the template matrix does also. If templateMatrixIdx missing: Completely dense matrix.
controls reuse of core variables instead of creation of new ones.
Note: only where template matrix indicates a nonzero.
If missing, all variables indicated by the use of the template matrix are created fresh.
same type for every newly created variable; default is 'C'; variables inherited from the core keep the type and bounds declared in the core.
must be constant matrix; zero if missing
contains varMtx - lbMtx; assumed all R^n if missing
must be constant matrix; zero if missing
contains ubMtx - varMtx; assumed all R^n if missing
number must match numberOfMatrixObj after "mult" has been considered
e.g., if symmetry = "upper", only objectives in the upper triangle (including diagonal) are generated and passed to the solver interface.
matrix objective has a nonzero entry wherever the template matrix does also. If templateMatrixIdx missing: Completely dense matrix.
controls reuse of core objectives instead of creation of new ones.
Note: only where template matrix indicates a nonzero.
If missing, all objectives indicated by the use of the template matrix are created fresh.
x is (weakly) preferred to y if obj(x) - obj(y) lies in cone. If reference uses only core objectives, orderConeIdx can be omitted and defaults to the corresponding orthant cone
added to objective; zero if missing
number must match numberOfMatrixCon after "mult" has been considered
number must agree with numberOfDescriptiors
can be used, e.g., to set up Toeplitz matrices
e.g., if symmetry = "upper", only variables in the upper triangle (including diagonal) are generated and passed to the solver interface.
matrix constraint has a nonzero entry wherever the template matrix does also. If templateMatrixIdx missing: Completely dense matrix.
controls reuse of core constraints instead of creation of new ones.
Note: only where template matrix indicates a nonzero.
If missing, all constraints indicated by the use of the template matrix are created fresh.
must be constant matrix; zero if missing
contains mtx expr - lbMtx; assumed all R^n if missing
must be constant matrix; zero if missing
contains ubMtx - mtx expr; assumed all R^n if missing
Number must agree with numberOfExpr
since version 2.0
refers to existing matrix objective (if negative) or matrix constraint (if nonnegative)
experiment--
For tensor-valued expressions only - scalar-valued expressions should be set up as nonlinearExpressions
number must match numberOfTensorVar after "mult" has been considered
Number of dimensions must equal the tensor rank
number must agree with numberOfDescriptiors
can be used to set up special tensors with properties not otherwise available
tensor variable has a nonzero entry wherever the template tensor does also. If templateTensorIdx missing: Completely dense tensor.
controls reuse of core variables instead of creation of new ones.
Note: only where template tensor indicates a nonzero.
If missing, all variables indicated by the use of the template tensor are created fresh.
same type for every newly created variable; default is 'C'; variables inherited from the core keep the type and bounds declared in the core.
must be constant tensor; zero if missing
contains varMtx - lbMtx; assumed all R^n if missing
must be constant tensor; zero if missing
contains ubMtx - varMtx; assumed all R^n if missing
number must match numberOfTensorObj after "mult" has been considered
Number of dimensions must equal the tensor rank
tensor objective has a nonzero entry wherever the template rtensor does also. If templateTensorIdx missing: Completely dense tensor.
controls reuse of core objectives instead of creation of new ones.
Note: only where template tensor indicates a nonzero.
If missing, all objectives indicated by the use of the template tensor are created fresh.
x is (weakly) preferred to y if obj(x) - obj(y) lies in cone. If reference uses only core objectives, orderConeIdx can be omitted and defaults to the corresponding orthant cone
added to objective; zero if missing
number must match numberOfTensorCon after "mult" has been considered
Number of dimensions must equal the tensor rank
number must agree with numberOfDescriptiors
can be used, e.g., to set up Toeplitz matrices
tensor constraint has a nonzero entry wherever the template tensor does also. If templateTensorIdx missing: Completely dense tensor.
controls reuse of core constraints instead of creation of new ones.
Note: only where template tensor indicates a nonzero.
If missing, all constraints indicated by the use of the template tensor are created fresh.
must be constant tensor; zero if missing
contains tensor expr - lbTensor; assumed all R^n if missing
must be constant tensor; zero if missing
contains ubMtx - mtx expr; assumed all R^n if missing
Number must agree with numberOfExpr
since version 2.0
refers to existing tensor objective (if negative) or tensor constraint (if nonnegative)
recommendation
Number of child elements must match numberOfSOS
recommendation
Only required if type >= 3. All nonzero variables must belong to the same adjacency set. Adjacency variable sets must be subsets of the SOS variable set.
If type = 2 adjacent variables are by default consecutive entries in the variables list.
"type = k" means at most k of the variables in this SOS can be nonzero, and the nonzeros must be "adjacent"
if present must refer to a previously defined constraint of the form
x_i1 + x_i2 + ... = 1.
experiment
Number of sets must match numberOfSets
since version 2.0
Number of various cone child elements must match numberOfCones
experiment
experiment. The cone of moments is the _dual_ of a corresponding cone of nonnegative polynomials
draft
use for tensors of rank >= 3.
draft
draft
draft
There must be (numberOfRows* NumberOfColumns) directions (after accounting for 'mult'). Row index changes fastest, then column index, then other indexes.
draft
draft
If present, the number of directions must correspond to the dimensions of the product Ax (after accounting for 'mult'). Row index changes fastest, then column index, then other indexes.
'positive': Ax \ge 0; 'negativ'e: Ax \le 0;
'mixed': specified by 'direction' element
draft
draft
This is an index of the form i0*n1*n2*... + i1*n2*n3... + ... + i_r, with zero-based indices i0, i1, ..., i_r
draft
Each is an index of the form i0*n1*n2*... + i1*n2*n3... + ... + i_r, with zero-based indices i0, i1, ..., i_r
draft
index is of the form i0*n1*n2*... + i1*n2*n3... + ... + i_r, with zero-based indices i0, i1, ..., i_r
draft
Each is an index of the form i0*n1*n2*... + i1*n2*n3... + ... + i_r, with zero-based indices i0, i1, ..., i_r
draft
draft
draft
dimension of direction vector must equal the number of variables in the polynomial
experiment. Used for both a cone of nonnegative polynomials and its dual, a moment cone
upper bounds on domain of polynomial. Default is +INF. Indexes are "flattened": i0*n1*n2*... + i1*n2*n3... + ... + i_r, with zero-based indices i0, i1, ..., i_r
lower bounds on domain of polynomial. Default is -INF. Indexes are "flattened": i0*n1*n2*... + i1*n2*n3... + ... + i_r, with zero-based indices i0, i1, ..., i_r
draft
each factor gives the index of a previously defined cone
draft
draft
each component gives the index of a previously defined cone
draft
draft
draft
experiment
recommendation
recommendation
number must agree with numberOfDataSources
finalization
if both uri and uriAlias missing, use defaultURI
if missing, use value of defaultRefreshPoint
number must agree with numberOfDataPoints
number must agree with numberOfTargetComponents
finalization
finalization
default is "real"
experiment
recommendation
recommendation
number must agree with numberOfDataItems
finalization
if both uri and uriAlias missing, use defaultURI
finalization
finalization
finalization
finalization
finalization
default is "real"
experiment
recommendation
recommendation
number must agree with numberOfDataItems
finalization
if both uri and uriAlias missing, use defaultURI
finalization
finalization
finalization
finalization
finalization
default is "real"
experiment
reference to an OSiL file containing the base for the current modifications
reference to a previous modification available in memory
defines a reference for subsequent reuse of this modification.
experiment
refers to a previous modification available in memory
experiment - refers to an osil file that may contain modifications from another base reference
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
delete a variable, its cost and constraint coefficients, and all quadratic terms and nonlinear expressions in which the variable occurs.
If present, delete only the objective coefficient in this column
delete a matrix variable and all terms in all nonlinear and matrix expressions in which the matrix variable occurs.
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
experiment
since version 1.0
since version 1.0
since version 1.0
experiment
experiment
experiment
experiment
experiment
proposal
Number of the disjunctiveSet child elements must agree with numberOfDisjunctiveSets
proposal
proposal
proposal
Number of the var child elements must agree with numberOfVariables
proposal
Number of the coef child elements must agree with numberOfObjCoef
proposal
Number of the obj child elements must agree with numberOfObjectives
proposal
proposal
Number of the con child elements must agree with numberOfConstraints
proposal
proposal
Number of the el child elements must agree with numberOfValues
if this location cooincdes with core information, the data get added together.
proposal
proposal
Number of the el child elements must agree with numberOfValues
if this location cooincdes with core information, the data get added together.
proposal
Number of the qTerm child elements must agree with numberOfQuadraticTerms
if this location cooincdes with core information, the data get added together.
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
Number of the disjunction child elements must agree with numberOfDisjunctions
proposal
proposal
These elements work at the model level (to include or exclude entire variables or constraints)
proposal
proposal
proposal
proposal
proposal
proposal
proposal
These elements affect the term level only (to include or exclude portions of a constraint or objective)
proposal
proposal
proposal
proposal
proposal
proposal
disjunction is enforced only if logical conditioning variable is TRUE
disjunction is enforced only if logical conditioning constraint is TRUE
proposal
proposal
Number of the el child elements must agree with numberOfValues
if this location coincides with core information, the data get added together.
proposal
proposal
Number of the el child elements must agree with numberOfValues
if this location coincides with core information, the data get added together.
proposal
Number of the qTerm child elements must agree with numberOfQuadraticTerms
if this location cooincdes with core information, the data get added together.
proposal
proposal
refers to an idx in the list of nonlinear expressions
proposal
proposal
proposal
proposal
Number of alternatives must agree with numberOfAlternatives
different disjunctive terms are connected by logical OR
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
Number of the var child elements must agree with numberOfVariables
if lb or ub are mentioned, they replace core information
proposal
Number of the obj child elements must agree with numberOfObjectives
if constant or weight is mentioned, it replaces information in the core
proposal
Number of the con child elements must agree with numberOfConstraints
if lb or up are mentioned, they replace core information
proposal
Number of the con child elements must agree with numberOfConstraints
proposal
Number of the con child elements must agree with numberOfConstraints
proposal
Number of the con child elements must agree with numberOfConstraints
proposal
Number of the con child elements must agree with numberOfConstraints
proposal
Number of disjunctions must agree with numberOfDisjunctions
to identify logical core variables that describe conditions under which one of the disjunctions holds
to identify core contraints that describe additional restrictions on the logical conditioningVariables
different disjunctive sets are connected by logical AND
experiment
disjunction is enforced only if logical conditioning variable is TRUE
disjunction is enforced only if logical conditioning constraint is TRUE
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
Number of the param child elements must agree with numberOfParameters
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
Number of the dist child elements must agree with numberOfDist
experiment
experiment
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
draft
draft
required if attr="coef"
draft
draft
draft
draft
draft
draft
draft
draft - may NOT change the sparsity pattern of the objective
draft
draft - may NOT change the sparsity pattern of the constraint matrix
draft - may NOT change the sparsity pattern of the quadratic expressions
draft - refer to any OSnLNode in any nonlinear expression, including matrix constructs
draft - may NOT change the sparsity pattern of the matrix
recommendation
recommendation
recommendation
recommendation
recommendation
recommendation
recommendation
recommendation
recommendation
recommendation
recommendation
recommendation
Number of the url child elements must agree with numberOfURIAliases
recommendation
recommendation
Number of the param child elements must agree with numberOfParameters
recommendation
recommendation
Number of the var child elements must agree with numberOfVariables
recommendation
recommendation
Number of the con child elements must agree with numberOfConstraints
recommendation
recommendation
Number of the obj child elements must agree with numberOfObjectives
recommendation
recommendation
Number of the el child elements must agree with numberOfValues
recommendation
Number of the qTerm child elements must agree with numberOfQuadraticTerms
recommendation
recommendation
recommendation
recommendation
Number of the number and string child elements must agree with numberOfNonlinearNumbersAndStrings
recommendation
recommendation
experiment
Number of the row child elements must agree with numberOfRows
experiment
proposal
experimental
experiment
Number of stages must agree with numberOfStages
proposal
proposal
proposal. If missing, assume ALL objectives are included
proposal. If missing, assume numberOfConstraints=0
proposal
List must be empty if startIdx is used and must match numberOfVar otherwise
proposal
proposal
List must be empty if startIdx is used and must match numberOfCon otherwise
proposal
proposal
List must be empty if startIdx is used and must match numberOfObj otherwise
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
Number of the scenario child elements (including the root scenario) must match numberOfScenarios
proposal
proposal
proposal
proposal
proposal
Number of the sNode child elements must match numberOfNodes
proposal
proposal
proposal
proposal
proposal
proposal
Number of row child elements must match numberOfRows
proposal
proposal
Number must match numberOfRandomFunctions
proposal
Number of distr child elements must match numberOfDistributions
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
Number must match numberOfNonlinearTransformations
proposal
proposal
proposal
Number must agree with numberOfStElem
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
number must match numberOfParameters
proposal
proposal
proposal
proposal
proposal
number must match numberOfParam
proposal
proposal
number must match numberOfValues
proposal
proposal
number must match numberOfValues
proposal
proposal
number must match numberOfValues
proposal
proposal
number must match numberOfValues
proposal
proposal
number must match numberOfValues
proposal
proposal
number must match numberOfValues
proposal
proposal
number must match numberOfValues
proposal
proposal
proposal
proposal
proposal
proposal
proposal
number of past nodes in the tree that enter the linking constraints
experiment
recommendation
number must agree with numberOfDataItems
finalization
number must agree with numberOfRandomElements
number must agree with numberOfTargetComponents
finalization
default is "real"
experiment
number must agree with numberOfDataItems
finalization
number must agree with numberOfDataPoints
number must agree with numberOfTargetComponents
if missing, assume identity matrix (error if dimension mismatch)
finalization
default is "real"
finalization
number must agree with numberOfARTerms
finalization
finalization
proposal
proposal
proposal
proposal
proposal
proposal
number of children must agree with numberOfStochasticElements
index refers to location within list of stochastic elements. Size of vector must agree with numberOfChanges
size of vector must agree with numberOfChanges
changes are relative to parent or, if present, the base scenario
proposal
Number of the sNode child elements must match numberOfNodes
proposal
proposal
governs the length of the history that can influence the present node as well as the offsets for variable and constraint references
proposal
proposal
proposal
experiment
Number of the nodes in this tree must match numberOfNodes
proposal
must be present if nodeData element is missing; if useCoreData is TRUE, ignore nodeData if present
proposal
governs the length of the history that can influence the present node as well as the offsets for variable and constraint references
proposal
used to provide a reference for later inheritance of data
proposal
since version 1.0
since version 1.0. The indexing is cumulative; current node indices are added after the indices of the variables in the linking ancestor nodes.
to add objective coefficients for variables in the currrent node into previously defined objectives
the indexing is cumulative. additional objectives are added below the indices of previously defined objectives.
since version 1.0. The indexing is cumulative; current node indices are added after the indices of the constraints in the linking ancestor nodes.
includes matrices on and below the diagonal. Variable indices are offset by the variables in the linking ancestor nodes
since version 1.0 - Does not use matrices. So, e.g. the expression 2x + 3y + x^2 + 2xy + 3y^2 has three quadratic coefficients: 1, 2, 3
since version 1.0. The indexing is cumulative; current node indices are added after the indices of the expressions in the linking ancestor nodes.
since version 2.0
We use the LPSOLVE definition of SOS, which goes back to a paper by J.A.Tomlin in "Studies on Graphs and Discrete Programming", 1981.
experiment
since version 2.0. The indexing is cumulative; current node indices are added after the indices of the ancestor nodes.
since version 2.0. The indexing is cumulative; current node indices are added after the indices of the ancestor nodes.
experiment
proposal
proposal
proposal
draft
experiment
experiment
experiment
finalization
finalization
realTimeData and simulations. experiment
experiment
finalization
experiment
used to identify another node in the tree as a base from which the current node inherits data
proposal
number must agree with numberOfBlocks.
The random variables referened by distributionIdx are independent from one block to another.
draft
proposal
default is "max" for minimization problems, "min" for maximization problems
number must agree with numberOfStochEl
number must agree with numberOfAR.
Each term adds rows to the input vector for the transformation, first the random variable or vector, followed by any references to stochEl.
number must agree with numberOfMA. Each moving average term refers to a previously set decision variable, matrix variable, etc. Each reference adds one or more rows to the input vector of the transformation
the transformation forms a flattened column vector of current observations and autoregressive terms and moving average terms (and perhaps a constant term) and transforms this into the vector of stochastic elements. Any random matrices are expanded column by column.
Default: identity matrix.
refers to a matrix in the "matrices" section.
number of the stage in which the information is revealed
gives the index of the distribution in the "distributions" element
autoregressive terms
moving average terms
gives the index of the distribution in the "distributions" element
experiment
number must agree with numberOfStochEl. Each index refers back to the list of stochEl in the first definition of this stochastic component
gives the index of a previously defined stochastic component ("stoch")
proposal
proposal
proposal
proposal
proposal
proposal
proposal
Number of the row child elements must match numberOfPenalties
proposal
proposal
proposal
proposal
proposal
proposal
proposal
proposal
Number of the probabilisticObjective child elements (simple or joint) must match numberOfProbabilisticObjectives
proposal
proposal
Number of the row child elements must match numberOfRows
proposal
Number of the chanceConstraint child elements (simple or joint) must match numberOfChanceConstraints
proposal
proposal
Number of the row child elements must match numberOfRows
proposal
Number of the integratedChanceConstraint child elements (simple or joint) must match numberOfIntegratedChanceConstraints
proposal
proposal
Number of the row child elements must match numberOfRows
proposal
Number of the userDefinedRiskMeasure child elements (simple or joint) must match numberOfUserDefinedRiskMeasures
proposal
proposal
Number of the row child elements must match numberOfRows
proposal
proposal
must agree with numberOfCon
proposal
proposal
number must agree with numberOfRows
experiment