System Modeler HTTP API Documentation

Version: 1.0.0

Resources

Application

Subresources

dataThe datasets loaded into the application environment.
workbookThe application environment's workbook.

GET

StringTypeValuesExplanationParameters
storedArray of StringsArray of StringsList of locally stored persisted environments.
informationArray of StringsArray of StringsInformation string associated with persisted environments.
eventsOutputSee documentationEvents that have occurred in the environment. If from is passed, these are from the specified event.
multiple getSee explanation.See explanation.Operational call to make multiple GET requests with a single request.

POST

StringExplanationSynchronicityParameters
persistPersist current environment.Synchronous
StringTypeValuesDefaultExplanation
filenameStringNon-empty stringsFilename to save current environment under.
filenameStringNon-empty stringsInformation string.
retrieveRetrieve persisted environment. Current environment will be discarded.Synchronous
StringTypeValuesDefaultExplanation
filenameStringNon-empty stringsFilename to load environment from.
removeDelete persisted environment.Synchronous
StringTypeValuesDefaultExplanation
filenameStringNon-empty stringsEnvironment to delete.

PATCH

No PATCH functionality.

Workbook

Subresources

The name of an existing project in the workbook.

GET

StringTypeValuesExplanationParameters
projectsArray of StringsArray of StringsThe names of existing projects in the workbook.

POST

StringExplanationSynchronicityParameters
newCreate a new project. If a dataset name is passed, the project will be created from that dataset. Otherwise it will be empty.Synchronous
StringTypeValuesDefaultExplanation
datasetStringStrings""(Optional) The dataset from which to create the new project.
deleteDeletes specified project.Synchronous
StringTypeValuesDefaultExplanation
projectStringStringsThe project to delete.
randomCreates a new random project.Asynchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee documentation

PATCH

No PATCH functionality.

Project

Subresources

modelsThe name of an existing model in the project.
utilitiesThe prototype utilities that will be attached to newly created models.
networkThe prototype network that will be passed to model learning or construction algorithms.
patternsThe set of models selected as patterns.

GET

StringTypeValuesExplanationParameters
nameStringNon-empty stringsName of the project.
typeStringNon-empty stringsModel type project is set to create.
modelsArray of StringsArray of non-empty stringsNames of models in project.
datasetStringStringsName of dataset attached to prototype network. Empty if no such dataset.
filterStringStringsName of filter applied to dataset attached to prototype network. Empty if no such dataset/filter.
ctestsArray of StringsArray of non-empty stringsThe names of available correlation tests.
dtestsArray of StringsArray of non-empty stringsThe names of available discretization tests.
ctest resultsOutputSee documentationThe results of available correlation tests.
dtest resultsOutputSee documentationThe results of available discretization tests.
possible variable typesArray of StringsArray of non-empty stringsThe types of variables permitted.
training topologyOutput or NULLNULL (without quotes), or as specified in output documentationSee documentation

POST

StringExplanationSynchronicityParameters
set typeSpecify the type of model the project is set to learn/create.Synchronous
StringTypeValuesDefaultExplanation
typeString"Multinomial", "Conditional Gaussian"The type of network to create from the project prototypes.
set structureSpecify a predefined structure to impose on the model the project is set to learn/create.Synchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Structure Input.
learnLearn a new model from the prototype network and utilities.Asynchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Learning Input.
k foldsPerform a hard K-Folds test (where the structure and parameters are learnt and then cross-validated) and then learn a new model from the prototype network and utilities.Asynchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Learning Input.
set dynamicsSet the dynamics for the model as a whole. To set dynamics on particular variables, call the appropriate function on the relevant nodes of the prototype network.Synchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Dynamics Input.
cloneClone (create duplicate of) the network.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of new project.
merge valuesMerge the values of the chosen variable. This will also result in the merging of raw values.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of variable.
valuesStringArray of StringsThe name of the variable whose values are to be merged.
add valuesAdd values to the chosen variable.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of variable.
valuesStringArray of StringsThe new values to be added. Note that value names must be unique.
remove valuesRemove values from the chosen variable. This will also remove all associated raw values.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of variable.
valuesStringArray of StringsThe values to be deleted. Associated raw values are also deleted.
rename valueRename a value of the chosen variable. Value names must be unique.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of variable.
valueStringNon-empty stringsCurrent name of value.
new nameStringNon-empty stringsNew name of value.
add variablesAdd variables to the prototype network. Adding and removing variables can only be done when the project has no models. However, projects can be cloned and then this can be performed on the new project.Synchronous
StringTypeValuesDefaultExplanation
namesStringArray of StringsThe names of the variables to be added.
remove variablesRemove variables from the prototype network. Adding and removing variables can only be done when the project has no models. However, projects can be cloned and then this can be performed on the new project.Synchronous
StringTypeValuesDefaultExplanation
namesStringArray of StringsThe names of the variables to be removed.

PATCH

StringExplanationSynchronicityParameters
nameRename project.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsNew name of project.
attached dataAttach dataset to project.Synchronous
StringTypeValuesDefaultExplanation
attached dataStringStringsName of dataset to attach to project's prototype network. An empty string will detach any current dataset. Should be placed before filter field if performing both. Filter defaults to "Unfiltered"
data filterAttach filter to project.Synchronous
StringTypeValuesDefaultExplanation
data filterStringStringsName of filter for attached dataset. An empty string will remove any current filters. Should be placed after dataset field if performing both.
priorPrior settings for the project.Synchronous
StringTypeValuesDefaultExplanation
priorInputSee documentationSee Discretization Input.

Model

Subresources

utilitiesThe utilies associated with the model.
masterThe model's master network.
resultsThe model's result sets.
A network of the model, indentified by index. To retrieve the master network, you should use the master resource path.

GET

StringTypeValuesExplanationParameters
nameStringNon-empty stringsName of model.
weight methodString"Score", "Equal", "Exponential", "Dynamic ML", "Dynamic MAP (Ave. Score)", "Dynamic MAP (Likelihood)"Specifications how network relative fitness/weight will be calculated.
networksIntegerPositive integer valuesNumber of networks.
datasetStringNon-empty stringsAttached dataset. Empty if none.
filterStringNon-empty stringsActive filter on attached dataset. Empty if no attached dataset.

POST

StringExplanationSynchronicityParameters
export syntheticUse the model to generate and export a synthetic dataset.Synchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Synthetic Dataset Input.
perform actionPerform an action on the model.Asynchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Action Input.
data caseSubmit a data case to be processed by the model.Synchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Data Feed Input.
add to modelAdd the networks of this model to another model.Synchronous
StringTypeValuesDefaultExplanation
modelStringNon-empty stringsName of model to add networks to.
add to patternsAdd this model to the project's pattern collection.Synchronous
StringTypeValuesDefaultExplanation

PATCH

StringExplanationSynchronicityParameters
nameRename model.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsNew name of model.
weightingSet how network relative fitness/weight will be calculated.Synchronous
StringTypeValuesDefaultExplanation
weightingStringNon-empty stringsSpecifies how network relative fitness/weight will be calculated.
dataSet attached dataset and filter. The filter is optional.Synchronous
StringTypeValuesDefaultExplanation
dataStringStringsName of dataset to attach.
filterStringStrings""Name of filter to place on attached dataset. (Optional)

Network

Subresources

Index of node.

GET

StringTypeValuesExplanationParameters
typeStringEnsemble, sub-network or uniqueNetwork type: 'ensemble' means that network is a overview of all networks in the model. 'subnetwork' means that it is a network in a model with multiple networks. 'unique' means that it is the network in a model with only one network.
fitnessArray of DoublesArray of double precision floating point valuesFitness of network, or of component networks if type is ensemble.
includedBooleantrue or falseWhether network will be included in actions on the model.
topologyOutputSee documentationTopological information about the network. The map's key values are the indices of the nodes. The value of the map is another map containing fields specifying the node's identity string (field 'id'), name (field 'name'), group (field 'group'), role (field 'role'), parent indices (field 'parents'), children indices (field 'children'), neighbor indices (field 'neighbors'), information parent indices (field 'iparents') and information child index (field 'ichild').

POST

StringExplanationSynchronicityParameters
toggle edgeSee Toggle Edge Input.Synchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee documentation
targeted climbSee Targeted Climb Input.Synchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee documentation

PATCH

StringExplanationSynchronicityParameters
includeWhether network should be included in actions on the model. Valid only for type==subnetwork networks.Synchronous
StringTypeValuesDefaultExplanation
includeBooleantrue or falseWhether network should be included in actions on the model. Valid only for type==subnetwork networks.

Node

Subresources

aprioriThe apriori probability distribution for the node.
aposterioriThe aposteriori probability distribution for the node. This is calculated from the last inference case.
relationshipsThe topological and informational relationships holding between the node and other nodes in the network, as well as information regarding compelled and prohibited relationships.
valuesThe values the variable associated with the node can take.
distributionThe conditional distribution associated with the node.

GET

StringTypeValuesExplanationParameters
nameStringStringsThe name of the node.
idIntegerNon-negative integer valuesThe id/index of the node.
roleStringStringsThe role of the node.
maximum parentsIntegerNon-negative integer valuesThe maximum number of parents of the node/variable.
mapped toStringStringsThe id/index of the prediction variable a mapped decision variable is mapped to. If there is no such variable, this returns "null"
value typeStringStringsThe value type of the node/variable.
dynamic offsetIntegerInteger valuesThe dynamic offset of the node.
groupStringStringsThe group the node belongs to.
datasetStringStringsThe dataset attached to the node. If no such data, string is empty.
data columnStringStringsThe data column attached to the node. If no such data, string is empty.
frequenciesOutputSee documentationFrequencies output providing information regarding the values found in attached data.
possible distributionsArray of StringsArray of StringsPossible distribution types.

POST

StringExplanationSynchronicityParameters
rediscretizeRediscretize the chosen node, using an algorithm specified by the passed input structure.Synchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Discretization Input.
set value typeSet the value type of the variable. Note that this can force a change in distribution type.Synchronous
StringTypeValuesDefaultExplanation
value typeStringNon-empty stringsDesired value type.
inputInputSee documentationSee Discretization Input.

PATCH

StringExplanationSynchronicityParameters
nameChange the name of the node.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of node. This can only be set for nodes in a project's prototype network and must be unique within the network.
roleChange the role of the node.Synchronous
StringTypeValuesDefaultExplanation
roleString"Chance", "Prediction", "Decision"Role of node in network.
maximum parentsChange the maximum number of parents of the node.Synchronous
StringTypeValuesDefaultExplanation
maximum parentsIntegerNon-negative integer valuesMaximum number of parents node can have.
mapped toChange id/index of the prediction variable a mapped decision variable is mapped to. -1 unsets mapped to.Synchronous
StringTypeValuesDefaultExplanation
mapped toIntegerInteger valuesIndex of the prediction node that a decision node is mapped to. The values of the variables associated with each node must share certain characteristics. This can only be set for nodes in a project's prototype network. -1 unsets current mapping.
groupChange the group the node is assigned to.Synchronous
StringTypeValuesDefaultExplanation
groupStringNon-empty stringsThe group the node is assigned to.
dataChange the data column assigned to the node. An empty string detaches the current data column.Synchronous
StringTypeValuesDefaultExplanation
dataStringNon-empty stringsThe data column to attach to the node. An empty string detaches the current data column.

Values

Subresources

No subresources.

GET

StringTypeValuesExplanationParameters
typeStringNon-empty stringsType of values variable has.
default intervalsIntegerPositive integer valuesDefault number of intervals for discretized variables. Note that this value is shared by ALL variables, so setting it for one sets it for all.
valuesArray of StringsArray of StringsValues of variable. If discrete, then all discrete values. If continuous, then empty.
raw valuesOutputSee documentationRaw values assigned to value of variable (mapping is by place in array). Raw values are strings / integers that are found 'raw' in datasets.
boundariesArray of DoublesArray of double precision floating point valuesInterval boundaries for discretized variables. If not discretized, then empty.

POST

No POST functionality.

PATCH

StringExplanationSynchronicityParameters
valueSet the raw values associated with a specific value of the variable.Synchronous
StringTypeValuesDefaultExplanation
valueStringNon-empty stringsValue whose associated raw values are to be set.
raw valuesStringArray of StringsRaw values to be set. Current raw values are overwritten.

Relationships

Subresources

No subresources.

GET

StringTypeValuesExplanationParameters
parentsArray of StringsArray of StringsGet list of parents of the node.
cparentsArray of StringsArray of StringsGet list of compelled parents of the node.
pcparentsArray of StringsArray of StringsGet list of possible (including actual) compelled parents of the node.
pparentsArray of StringsArray of StringsGet list of prohibited parents of the node.
ppparentsArray of StringsArray of StringsGet list of possible (including actual) prohibited parents of the node.
iparentsArray of StringsArray of StringsGet list of information parents of the node.
piparentsArray of StringsArray of StringsGet list of possible (including actual) information parents of the node.
iparentsArray of StringsArray of StringsGet list of information parents of the node.
childrenArray of StringsArray of StringsGet list of children of the node.
cchildrenArray of StringsArray of StringsGet list of compelled children of the node.
pcchildrenArray of StringsArray of StringsGet list of possible (including actual) compelled children of the node.
pchildrenArray of StringsArray of StringsGet list of prohibited children of the node.
ppchildrenArray of StringsArray of StringsGet list of possible (including actual) prohibited children of the node.
ichildStringStringsGet information child of the node. String is empty if no such node.
pichildrenArray of StringsArray of StringsGet possible (including actual) information children of the node.

POST

No POST functionality.

PATCH

StringExplanationSynchronicityParameters
cparentsSet compelled parents of the node.Synchronous
StringTypeValuesDefaultExplanation
cparentsStringArray of StringsList of nodes to be set as the compelled parents of the node.
pparentsSet prohibited parents of the node.Synchronous
StringTypeValuesDefaultExplanation
pparentsStringArray of StringsList of nodes to be set as the prohibited parents of the node.
iparentsSet information parents of the node.Synchronous
StringTypeValuesDefaultExplanation
iparentsStringArray of StringsList of nodes to be set as the information parents of the node.
cchildrenSet compelled children of the node.Synchronous
StringTypeValuesDefaultExplanation
cchildrenStringArray of StringsList of nodes to be set as the compelled children of the node.
pchildrenSet prohibited children of the node.Synchronous
StringTypeValuesDefaultExplanation
pchildrenStringArray of StringsList of nodes to be set as the prohibited children of the node.
ichildSet information child of the node. An empty string unsets current information child.Synchronous
StringTypeValuesDefaultExplanation
ichildStringStringsName of the node to be set as the information child of the node. An empty string unsets current information child.

Probabilities

Subresources

No subresources.

GET

StringTypeValuesExplanationParameters
typeString"Multinomial", "Conditional Gaussian", "Delta"The type of distribution.
parametersOutputSee documentationThe parameters of the distribution. For multinomial distributions, this is an array giving the Dirichlet hyper-parameters. For Gaussian mixture distributions, this is an array of arrays giving the (i) weight (ii) mean, and (iii) variance of each Gaussian.

POST

No POST functionality.

PATCH

No PATCH functionality.

Distribution

Subresources

No subresources.

GET

StringTypeValuesExplanationParameters
typeString"Multinomial", "Conditional Gaussian", "Delta"The type of distribution.
customBooleantrue or falseWhether the distribution has custom (initial) parameters.
adaptableBooleantrue or falseWhether the parameters are adaptable. They can only be non-adaptable if they are also custom.
parametersOutputSee documentationThe type specific parameters for the distribution.

POST

No POST functionality.

PATCH

StringExplanationSynchronicityParameters
typeSet distribution type.Synchronous
StringTypeValuesDefaultExplanation
typeStringNon-empty stringsThe type of distribution.
customSpecify if distribution has custom parameters.Synchronous
StringTypeValuesDefaultExplanation
customBooleantrue or falseWhether distribution has custom parameters.
adaptableSpecify if distribution has adaptable custom parameters.Synchronous
StringTypeValuesDefaultExplanation
adaptableBooleantrue or falseWhether distribution has adaptable custom parameters.
parametersType specific specification of parameters.Synchronous
StringTypeValuesDefaultExplanation
parametersInputSee documentationSee documentation

Results

Subresources

The index of an existing action result set, or"feed" for datafeed.

GET

StringTypeValuesExplanationParameters
actionsArray of StringsArray of StringsList of action result sets present, excluding that associated with the datafeed.
sizesArray of IntegersArray of IntegersSize of each action result set present, excluding that associated with the datafeed.
decisionsArray of Arrays of StringsArray of StringsList of decision variables for each action result set present, excluding that associated with the datafeed.
mappedArray of Arrays of BooleansArray of StringsSpecification of whether decision variables are mapped to prediction variables for each action results, excluding that associated with the datafeed.
predictionsArray of Arrays of StringsArray of StringsList of prediction variables for each action result set present, excluding that associated with the datafeed.

POST

StringExplanationSynchronicityParameters
deleteDelete specified action result set.Synchronous
StringTypeValuesDefaultExplanation
indexIntegerNon-negative integer valuesIndex of action result set to delete.
clearDelete all action result sets. Not that this does not affect the datafeed result set.Synchronous
StringTypeValuesDefaultExplanation
reset datafeedClear the datafeed result set.Synchronous
StringTypeValuesDefaultExplanation

PATCH

No PATCH functionality.

Result Set

Subresources

No subresources.

GET

StringTypeValuesExplanationParameters
itemsIntegerPositive integer valuesNumber of items in result set.
predictionsArray of StringsArray of non-empty stringsPrediction variables.
decisionsArray of StringsArray of non-empty stringsDecision variables.
mappedArray of BooleansArray of BooleansWhether respective decision variable is mapped to a prediction variable.
retrieveOutputSee documentationFetch a particular record or set of records from the result set.
get confusionOutputSee documentationFetch a particular confusion matrix from the result set.

POST

StringExplanationSynchronicityParameters
exportExport a particular record or set of records from the result set.Asynchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee documentation

PATCH

No PATCH functionality.

Utilities

Subresources

Index of the utility function to retrieve.

GET

StringTypeValuesExplanationParameters
sizeIntegerNon-negative integer valuesNumber of utility functions in collection.
functionsSee explanation.See explanation.Operational call to make GET requests on all utility functions with a single request. Returns array of responses.

POST

StringExplanationSynchronicityParameters
newCreate a new, empty utility function.Synchronous
StringTypeValuesDefaultExplanation
eraseDelete the specified utility function.Synchronous
StringTypeValuesDefaultExplanation
indexIntegerNon-negative integer valuesThe index of the utility function to delete.

PATCH

No PATCH functionality.

Utility Function

Subresources

No subresources.

GET

StringTypeValuesExplanationParameters
variablesOutputSee documentationAn array specifying variable specific settings. Each item is a map. Fields are: (i) "name", the variable name. (ii) "values", which specifies the number of discrete values the variable has. (iii) "continuous", which specifies if the variable is to be treated as continuous (for continuous and disccretized variables only), and with values 'true' or 'false'. True will take the given continuous value as argument for the equation. (iv)"exponent", which, for variables where "continuous" is set to true, specifies if the value should be the exponent of the variable value. (v) "custom value", which specifies the custom values to be passed to the utility function when the variable takes the appropriate value (specified by order).
equationStringStringsA specification of the last applied equation. This may not reflect current values if they have been manually edited after equation application.
valuesArray of DoublesArray of double precision floating point valuesWhen all variables are discrete, this specifies the values in the utility potential.

POST

StringExplanationSynchronicityParameters
set variablesSet variables involved as arguments in the utility function.Synchronous
StringTypeValuesDefaultExplanation
variablesStringArray of StringsNames of the variables involved as arguments in the utility function.
apply equationApply a specified equation.Synchronous
StringTypeValuesDefaultExplanation
variablesStringStringsThe equation associated to be applied to the utility function.
edit variablesSet variable specific settings.Synchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee documentation
set valuesSet specific values in discrete variable only functions.Synchronous
StringTypeValuesDefaultExplanation
indicesArray of DoublesArray of double precision floating point valuesThe indices of the parameters to be set.
valuesArray of DoublesArray of double precision floating point valuesThe values to set the parameters to. Must be ordered to correspond with the passed indices.

PATCH

No PATCH functionality.

Patterns

Subresources

resultsThe result set collection for actions and live feeds using this pattern collection object.

GET

StringTypeValuesExplanationParameters
patternsArray of StringsArray of StringsModels included as patterns in the pattern collection.

POST

StringExplanationSynchronicityParameters
add modelAdd model to the pattern collection.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of model to add to pattern collection.
remove modelRemove model from the pattern collection.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of model to remove from pattern collection.
evaluatePerform an action on the pattern collection.Asynchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Evaluate Patterns Input.

PATCH

No PATCH functionality.

Database

Subresources

The names of dataset present in environment.

GET

StringTypeValuesExplanationParameters
datasetsArray of StringsArray of StringsNames of datasets present in environment.
columnsArray of Arrays of StringsArray of Arrays of StringsNames of datasets present in environment.
filtersArray of Arrays of StringsArray of Arrays of StringsNames of datasets present in environment.

POST

StringExplanationSynchronicityParameters
deleteDelete the specified dataset.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsName of the dataset to be deleted.
import odbcImport a dataset from an ODBC source.Asynchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Import SQL Input.
import textImport a dataset from a text source.Asynchronous
StringTypeValuesDefaultExplanation
inputInputSee documentationSee Import Text Input.

PATCH

No PATCH functionality.

Dataset

Subresources

No subresources.

GET

StringTypeValuesExplanationParameters
nameStringStringsName of the dataset.
columnsArray of StringsArray of StringsThe names of the columns in the dataset.
rowsIntegerInteger valuesThe number of rows in the dataset.
filtersArray of StringsArray of StringsThe filters associated with the dataset.
bmvsArray of StringsArray of StringsThe binary meta-variable columns of the dataset.
metaArray of StringsArray of StringsThe meta-information columns of the dataset.
dataOutputSee documentationThe data contained in the dataset, returned as an array of arrays of strings. If from and to values are passed, then these represent the first row return and the first row not returned.

POST

StringExplanationSynchronicityParameters
deleteDelete column from dataset.Synchronous
StringTypeValuesDefaultExplanation
columnStringNon-empty stringsName of column to be deleted.
renameRename column in dataset. Note that column names must be unique.Synchronous
StringTypeValuesDefaultExplanation
columnStringNon-empty stringsCurrent name of column to be renamed.
new nameStringNon-empty stringsNew name of column.
add to bmvAdd column to a binary meta-variable (BMV) column. If the BMV column is not specified, a new BMV column is created.Synchronous
StringTypeValuesDefaultExplanation
columnStringNon-empty stringsName of column to be add to the BMV.
new nameStringStrings"create new"Name of BMV column. (Optional: If the BMV column is not specified, this defaults to "Create New" which causes a new BMV column to be created.)
set as metaSet/unset column as meta information.Synchronous
StringTypeValuesDefaultExplanation
columnStringNon-empty stringsName of column.
metaBooleantrue or falseWhether column contains meta information.
exportExport dataset to an ODBC source, as the specified table.Synchronous
StringTypeValuesDefaultExplanation
columnStringNon-empty stringsAn unused table name in the ODBC source. The table will be generated and the dataset exported to it.
edit filtersAdd and remove filters.Synchronous
StringTypeValuesDefaultExplanation
taskString"add" or "remove"Task to perform.
filterStringNon-empty stringsFilter encoding / name.

PATCH

StringExplanationSynchronicityParameters
nameRename the specified dataset.Synchronous
StringTypeValuesDefaultExplanation
nameStringNon-empty stringsNew name of the dataset to be rename.

Input

Input Base

FieldTypeValuesOptionalDefaultExplanation
contextStringCannot be empty or white space.truedefaultSpecifies the context the current action will be taken in. Using custom contexts permits targeted cancellation of tasks and error reporting. Note that through it's inclusion in the Input_Base object, this field is present in ALL inputs.

Data Generation Input

FieldTypeValuesOptionalDefaultExplanation
missing data algorithmString"Gibbs Sampler", "Expectation Maximization", "Ignore", "Discard"trueIgnore
gibbs burnIntegerNon-negative integertrue100Number of samples to discard at the beginning of the GS algorithm.
gibbs samplesIntegerPositive integertrue1000Number of samples to include (in addition to the burn) in the GS algorithm.
em candidatesIntegerPositive integertrue1Number of candidate initial values to include in the EM algorithm.
em divisionsIntegerPositive integertrue100Maximum number of divisions to make on a single datum in the EM algorithm.
em divisionsIntegerPositive integertrue100Maximum number of divisions to make on a single datum in the EM algorithm.
em preferenceBoolean'true','false'truetrueWhether the EM algorithm will be replaced by the GS algorithm when the divisions limit is breached. If false, the offending datum is discarded.
score metricString"Bayesian Dirichlet Score", "Maximum Likelihood", "Maximum Likelihood (AIC penalty)", "Maximum Likelihood (BIC penalty)", "Weka Entropic Metric (AIC penalty)", "Weka Entropic Metric (BIC penalty)"trueBayesian Dirichlet ScoreThe scoring metric to be used in the learning algorithm.

A Priori Input

FieldTypeValuesOptionalDefaultExplanation
performString'true','false'truetrueWhether a priori inference should be performed at the completion of model construction/learning.
inference algorithmString"Junction Tree", "Variable Elimination", "Importance Sampling"trueJunction TreeInference algorithm to use.
sample sizeIntegerPositive integertrue1000Sample size of sampling inference algorithms.

Learning Input

FieldTypeValuesOptionalDefaultExplanation
modelStringUnique in project scope.false-Name of model to be learnt.
search spaceString"Mixed Equivalence Class Search", "Unordered DAG Equivalence Classes ordered by Inclusion", "Unordered DAG Equivalence Classes", "Unordered DAGs", "Ordered DAGs"trueMixed Equivalence Class SearchSpace of possible structures to search.
min improvementDouble precision floating point.Non-negative integertrue0Minimum accepted improvement in model fitness for an iteration of the search algorithm.
ensemble methodString"Unique", "Average", "Compound"trueUniqueMethod of combining networks.
sampling methodString"Restart", "Bootstrap"trueRestartSpecifies the type of difference that occurs in different iterations of the learning algorithm.
ensemble weightingString"Score", "Equal", "Exponential", "Dynamic ML", "Dynamic MAP (Ave. Score)", "Dynamic MAP (Likelihood)"trueScoreHow networks in an ensemble model will be weighted.
min weightDouble precision floating point.Non-negative integertrue0A parameter controlling the relative fitness a network must have to be included in an ensemble. Fitness is relative to current best network.The parameter functions as follows: Let this parameter be p. Remembering that score metric are log values (and hence negatives), if model A is the current best model and has fitness n, model B with weight m will be accepted only if m-(n*p)>=n.
max modelsIntegerPositive integertrue5Maximum number of networks in the ensemble model.
self convergencesIntegerNon-negative integertrue10Number of times self-convergence (looping) is permitted in an iteration of the learning algorithm. Self convergence can occur when working with missing data.
learning algorithmString"Hill Climb"trueHill ClimbLearning algorithm to be used.
edgeDouble precision floating point.[0,1)true.05Probability of an edge being present in a random restart.
target predictionsBoolean'true','false'truefalseIf true, only edges connected to a prediction variable are involved in generating a random restart's initial structure.
persistent edgesBoolean'true','false'truefalseWhether edges added for a random restart should be unable to be removed during the learning algorithm.
terminationSee explanationOne or more of:max restarts, max non convergence, max non selection, max timetruemax restartsConditions under which the learning algorithm should terminate.
max restartsIntegerNon-negative integertrue0Maximum number of restarts before termination. Note that it is a non-negative number since entering '0' will produce a single start and zero restarts.
max non convergenceIntegerPositive integertrue0Maximum number of non-converging iterations of the learning algorithm before termination.
max non selectionIntegerPositive integertrue0Maximum number of non selected iterations of the learning algorithm before termination.
max timeDouble precision floating point.Positive integertrue0Time before termination,
apriori settingsSee A Priori Input.See A Priori Input.trueSee A Priori Input.Specifications of desired a priori behavior.
learning data settingsSee Data Generation Input.See Data Generation Input.trueSee A Priori Input.Specifications of desired learning data behavior.

Import SQL Input

FieldTypeValuesOptionalDefaultExplanation
selectStringA valid SQL select query.false-A valid SQL select query.

Import Text Input

FieldTypeValuesOptionalDefaultExplanation
missingStringNon-empty Stringsfalse-Specifications of strings that represent missing values. Note that these apply to all variables. If a valid string for one variable represents a missing value for another, you should edit the text file prior to import.
typesArray of StringsNon-empty Stringsfalse-Specifications of variable types.
dividersArray of charactersCharacter values.false-Specifications of dividers between variable values. Note that these are characters. If your dividers are multi-character strings, you should edit the text file prior to import.
true stringsArray of StringsNon-empty StringstrueNo default value, but can be left empty if no Boolean values.Specifications of the strings that represent true for particular variables. You can either enter no such values, or values for all variables, passing empty strings for non-boolean variables.
false stringsArray of StringsNon-empty StringstrueNo default value, but can be left empty if no Boolean values.Specifications of the strings that represent false for particular variables. You can either enter no such values, or values for all variables, passing empty strings for non-boolean variables.
namesBoolean'true','false'truefalseWhether the first row of the text contains the names of the variables.
reserveIntegerPositive integertrue10000Estimate of the number of rows in the file. Passing an optimistic but reasonable value can considerably speed up large imports.

Discretization Input

FieldTypeValuesOptionalDefaultExplanation
algorithmString"Equinumeric", "Equilength", "StdDev", "Custom_Length", "K_Means"trueEquinumericDiscretization algorithm to use.
fromDouble precision floating point.Double precision floating point valuestrue0From value.
toDouble precision floating point.Double precision floating point valuestrue1To value.
intervalsIntegerPositive integertrue8Number of intervals to create.
kmeans startString"Random", "Equidistant"trueEquidistantHow the initial assignment of values should be chosen for the K-Means discretization algorithm.
std devBoolean'true','false'truefalseWhether empty intervals should be permitted in the K-Means discretization algorithm.
kmeans permits zeroIntegerPositive integertrue200The maximum number of iterations before termination for the K-Means algorithm.
kmeans safety limitDouble precision floating point.Double precision floating point valuestrue3The range in terms of standard deviations to create intervals in for the standard deviation discretization algorithm.
insertBooleanBooleantruefalseWhether the newly created intervals should be inserted into the currently existing intervals. Only valid when working with domain variables (not model variables).

Action Input

FieldTypeValuesOptionalDefaultExplanation
nameStringUnique in model scope.trueUnknownName of action.
typeString"Actual", "Test", "Ad-Hoc", "K-Folds"trueActualType of action to perform.
adaptionString"No Adaption"trueNo AdaptionWhether and how models should adapt to the data involved in the action.
inferenceString"Junction Tree", "Variable Elimination", "Importance Sampling"trueJunction TreeInference algorithm to use in action.
samplesIntegerPositive integertrue1000Number of samples to be taken for sampling inference algorithms.
decisionTestString"basic", "prediction", "expected", "both"truebothWhether a decision test action should also perform pure predictive tests and/or calculate expected utility without intervention.
networkIntegerNon-negative integertrue0The network to perform the action on.

Cross Validation Input

FieldTypeValuesOptionalDefaultExplanation
algorithmString"Intervals", "Values"trueIntervalsCross validation algorithm to use.
foldsIntegerPositive integertrue10Number of sub-sets to divide data into for interval based cross validation.
indexIntegerNon-negative integertrue0Index of variable to use for values based cross validation.
randomBoolean'true','false'truefalseWhether data should be randomized.
adaptionString"No Adaption"trueNo AdaptionAdaption algorithm to use.
inferenceString"Junction Tree", "Variable Elimination", "Importance Sampling"trueJunction TreeInference algorithm to use.
samplesIntegerPositive integertrue1000Number of samples to use in sampling inference algorithms, if applicable.
keep modelsBoolean'true','false'truefalseWhether the models built from each cross validation should be kept.

Ad-hoc Data Input

FieldTypeValuesOptionalDefaultExplanation
discreteArray of IntegersNon negative Integersfalse-Integer values for discrete variables. UNKNOWN_VALUE (INT_MIN) represents unknown values. If all variables are continuous, this need not be passed. Otherwise, all variables must have an entry but entries for continuous variables will be ignored. At least one of the continuous and discrete fields must be passed.
continuousArray of double precision floating pointDouble precision floating point valuesfalse-Double values for discrete variables. UNKNOWN_CONTINUOUS_VALUE (-DBL_MAX) represents unknown values. If all variables are discrete, this need not be passed. Otherwise, all variables must have an entry but entries for discrete variables will be ignored. At least one of the continuous and discrete fields must be passed.

Data Feed Input

FieldTypeValuesOptionalDefaultExplanation
settingsSee Action Input.See Action Input.false-See Action Input.
dataSee Ad-hoc Data Input.See Ad-hoc Data Input.false-See Ad-hoc Data Input.

Targeted Climb Input

FieldTypeValuesOptionalDefaultExplanation
intervalsIntegerNon-negative integertrue10Number of alternative interval numbers to check on discretized variables. Note this is +/- current number for each variable.
targetIntegerNon-negative integerfalse-ID of variable to target.
cross validationSee Cross Validation Input.See Cross Validation Input.false-See Cross Validation Input.
learningSee Learning Input.See Learning Input.false-See Learning Input.
discretization settingsArray of Discretization Inputs. See Discretization Input.See Discretization Input.false-See Discretization Input.

Toggle Edge Input

FieldTypeValuesOptionalDefaultExplanation
fromIntegerNon-negative integerfalse-The ID of the node the edge comes from.
toIntegerNon-negative integerfalse-The ID of the node the edge goes to.
learn data settingsSee Data Generation Input.See Data Generation Input.false-See Data Generation Input.

Synthetic Dataset Input

FieldTypeValuesOptionalDefaultExplanation
sizeIntegerPositive integerfalse-The number of rows to be generated.
unknownArray of double precision floating point[0,1)false-The chance each variable is unknown.

Hard K-Folds Input

FieldTypeValuesOptionalDefaultExplanation
cross validationSee Cross Validation Input.See Cross Validation Input.false-See Cross Validation Input.
learningSee Learning Input.See Learning Input.false-See Learning Input.

Evaluate Patterns Input

FieldTypeValuesOptionalDefaultExplanation
dataStringCannot be empty or white space.false-Pattern which has attached the data that will be used for the action.
actionType of action to perform."Actual", "Test", "Ad-Hoc", "K-Folds"trueTestType of action to perform.
inferenceType of inference algorithm to use."Junction Tree", "Variable Elimination", "Importance Sampling"trueJunction TreeType of inference algorithm to use.
nameName of action.Cannot be empty or white space.trueunknownName of action.

Random Input

FieldTypeValuesOptionalDefaultExplanation
domainStringCannot be empty or white space.truerandom_domainName of new domain. If no string is passed the default name will be "random_domain" plus, if required, an appropriate index to ensure that it is unique within environment scope.
modelStringCannot be empty or white space.truemodelName of new model.
variablesIntegerPositive integertrue0Number of variables.
edgeDouble precision floating point.[0,1)true.1Probability of an edge between two nodes.
valuesIntegerPositive integertrue4Maximum or constant number of values. Must be 2 or higher.
parentsIntegerPositive integertrue4Maximum number of parents
constantBoolean'true','false'truefalseWhether the values field specifies the number of values all variables have (true) or the maximum number of values a variable can have (false). If the second, variables are assigned a random number of values between 2 and the specified limit.
aprioriSee A Priori Input.See A Priori Input.trueSee A Priori Input.See A Priori Input.

Structure Input

FieldTypeValuesOptionalDefaultExplanation
settingString"Naive Bayes", "Soft Naive Bayes", "Temporal", "Clear"trueNaive BayesPre-specified structural constraint to place on domain.
targetStringCannot be empty or white space.false-Name of target prediction variable.

Dynamics Input

FieldTypeValuesOptionalDefaultExplanation
dynamicsArray of IntegersIntegerstrueEmpty Json Map.Json Map from variable name to desired dynamics specified as Json Arrays of integers. The integers specify the dynamic offsets involved. Absence of variable indicates no dynamics specified for that variable.
targetStringString.trueWhether dynamics is aimed at a particular variable. Otherwise the function will edit all variables.
overrideBoolean'true','false'truefalseWhether dynamics should be added to current dynamics (false) or override current dynamics (true). To clear dynamics, enter "true" with empty or default dynamicsfield.

Export Results Input

FieldTypeValuesOptionalDefaultExplanation
tableStringString.false-ODBC table to export to. If it does not exist, it will be created.
firstIntegerPositive integertrue1First item to be exported. Default is 1 meaning all items from the first will be exported. Note that the items are indexed from 1, not 0.
lastIntegerPositive integertrue-1Last item to be exported. Default is -11 meaning all items from the specified first item will be exported. Note that the item gves the last item exported, not the first not exported.

Prediction Filter Input

FieldTypeValuesOptionalDefaultExplanation
codeStringCannot be empty or white space.trueUnique PredictionThe type of macro output to return. Note that all such outputs can be deduced by client from other information given.
use_confidenceBoolean'true','false'truefalseWhether confidence estimations should be used in calculating doubt. If so, the lower confidence interval (at the confidence specified in the confidence_hurdle field) must be above the specified doubt hurdle.
confidence_hurdleDouble precision floating point.[0,1)true0.05The total percentage of the density that is to be outside both upper and lower confidence intervals. So half this amount will be above the upper 'error bar' or below the lower 'error bar'.
doubt_hurdleDouble precision floating point.[0,1)true0The minimum probability level required for the MAP value to not be classified as doubt. Note that in the case of a doubt classification, the actual MAP value can still be recovered from passed information.
density_hurdleDouble precision floating point.[0,1)true0.95The percentage of the density to be included in a range or set macro classification.
confidence_undefinedBoolean'true','false'truetrueWhether a prediction should NOT be classified as doubt when confidence estimation fails.
orderStringCannot be empty or white space.trueUnorderedThe type of ordering that holds on the variable's values.