2215 questions found.

Page 45 / 45

Yes, the 'maxWeight' parameter allows you to specify a maximum weight. The action will only return cuts whose total link weight is less than or equal to this value....
Read full answer See technical action

The minCut action can produce two main output tables: 'outCutSets' to store the links that form the minimum cut sets, and 'outPartitions' to store the minimum cut partitions of the nodes....
Read full answer See technical action

Yes, you can specify whether the graph should be treated as directed or undirected using the 'direction' parameter, which defaults to 'UNDIRECTED'....
Read full answer See technical action

The minSpanTree action is used to calculate the minimum spanning tree of a graph. A minimum spanning tree is a subgraph that connects all the vertices together with the minimum possible total edge weight, without forming any cycles....
Read full answer See technical action

The action supports both directed and undirected graphs. You can specify the graph type using the 'direction' parameter. For directed graphs, a 'source' node can be specified to find a spanning forest rooted at that node....
Read full answer See technical action

The primary inputs are the 'links' table, which defines the connections and weights between nodes, and optionally a 'nodes' table for node-specific information. Alternatively, you can use a pre-existing in-memory graph specified by the 'graph' parameter....
Read full answer See technical action

The main output is a data table, specified by the 'out' parameter, which contains the links that constitute the minimum spanning tree. The total weight of the resulting tree is provided in the 'objective' field of the results....
Read full answer See technical action

Yes, the action supports parallel processing. You can use the 'nThreads' parameter to specify the number of threads for multithreaded processing, and the 'distributed' parameter to run the algorithm on a distributed graph for large-scale problems....
Read full answer See technical action

The mitigateBias action, part of the Fair AI Tools Action Set, is used to mitigate bias during the training of predictive models using the exponentiated gradient reduction algorithm....
Read full answer See technical action

The `biasMetric` parameter allows you to specify one of four types of bias measurements: 'DEMOGRAPHICPARITY', 'EQUALIZEDODDS', 'EQUALOPPORTUNITY', or 'PREDICTIVEPARITY'. The default is 'PREDICTIVEPARITY'....
Read full answer See technical action

The required parameters are `table` to specify the input data, `response` to specify the target variable, `sensitiveVariable` to define the variable for bias calculation, and `trainProgram` which contains the CASL code for model training....
Read full answer See technical action

The `trainProgram` parameter specifies the CASL training code to be executed. The mitigation action passes several key variables to this program, such as the modified input table (`tableCASLVariable`), the weight variable (`weightCASLVariable`), and the current iteration number (`iterationCASLVariab...
Read full answer See technical action

The `tolerance` parameter specifies the parity constraint violation tolerance. The mitigation process stops when the bias measurement falls below this tolerance. If set to 0, the action will run for the maximum number of iterations specified by `maxIters`....
Read full answer See technical action

You can use the `maxIters` parameter to set the maximum number of iterations for the exponentiated gradient reduction algorithm (default is 10) and the `learningRate` parameter to define the step size for updates (default is 0.01)....
Read full answer See technical action

Yes, the `tableSaveList` parameter allows you to specify a list of tables to save after running the training program in an iteration. These tables are saved only if the specified `biasMetric` improves during that iteration....
Read full answer See technical action
Copyright © WeAreCAS 2014-2026
SAS and all other SAS Institute Inc. product or service names are registered trademarks or trademarks of SAS Institute Inc. in the USA and other countries. ® indicates USA registration. WeAreCAS is an independent community site and is not affiliated with SAS Institute Inc.