Frequently Asked Questions

550 questions found.

Page 4 / 11

The `scoringAlgorithm` parameter allows you to choose between two algorithms: 'FREQUENCY' (which is the default) and 'WEIGHTED'....
Read full answer See technical action

The `casOut` parameter specifies the main output data table for the categories. The `matchOut` parameter creates a table with one match per row. The `groupedMatchOut` parameter creates a table that groups all matches for a given category within each document, using the delimiter specified by `matchD...
Read full answer See technical action

The `docId` parameter is used to specify the name of the variable in the input table that serves as a unique identifier for each document....
Read full answer See technical action

The `applyConcept` action performs concept extraction on a given text using a concept extraction model, which is defined in a LITI (Language Interpretation for Text Information) file....
Read full answer See technical action

The primary input is a CAS table specified by the `table` parameter, containing the documents to process. The `docId` and `text` parameters specify the columns for the document ID and the text content, respectively. The main output is the `casOut` table, which contains the concept match results. Oth...
Read full answer See technical action

You can specify a user-defined LITI model using the `model` parameter. This parameter points to a CAS table that contains your custom model. If this parameter is not specified, the base model is used....
Read full answer See technical action

The `matchType` parameter specifies the matching strategy for concepts. It can be set to 'ALL' (default) to return all matches, 'BEST' to return only the best-scoring matches, or 'LONGEST' to return the longest matches....
Read full answer See technical action

To improve performance, especially when using the CLAUS_n operator, you can use a two-step process. First, run `applyConcept` with the `parseTableOut` parameter to save the pre-parsed documents to a CAS table. Then, in a subsequent run, use this table as input via the `parseTableIn` parameter to avo...
Read full answer See technical action

Yes, you can use the `dropConcepts` parameter to provide a list of concept names that you want to exclude from the output tables. This is useful for filtering out predefined or intermediate concepts from the final results without removing them from the model itself....
Read full answer See technical action

The `litiChunkSize` parameter specifies the size of the data chunks used for processing a document. The default is '32K'. For very large documents, using a smaller chunk size like '32K' or '64K' can improve performance and reduce memory consumption. Setting it to 'ALL' processes the entire document ...
Read full answer See technical action

The arima action provides functionality for Autoregressive Integrated Moving Average (ARIMA) models, which are used for analyzing and forecasting univariate time series data....
Read full answer See technical action

The autoregressive polynomials are specified using the 'p' parameter within the 'estimate' options. You can define one or more polynomial factors, each with its list of lags....
Read full answer See technical action

The 'q' parameter is used to specify the moving average (MA) polynomials of the ARIMA model, similar to how the 'p' parameter specifies the autoregressive (AR) part....
Read full answer See technical action

The 'method' parameter allows you to choose the estimation method. The available options are CLS (Conditional Least Squares), ML (Maximum Likelihood), and ULS (Unconditional Least Squares)....
Read full answer See technical action

You can specify the differencing orders using the 'diff' parameter within the 'estimate' options. This parameter accepts a list of integers representing the differencing lags....
Read full answer See technical action

The 'transform' parameter can apply a transformation to the time series. The available options are AUTO, BOXCOX (with a parameter specified by 'transformParm'), LOG (logarithmic), LOGIT (logit), SQRT (square root), and NONE (no transformation)....
Read full answer See technical action

You can generate forecasts by specifying the 'forecast' option within a 'model' definition. Key parameters include 'lead' to set the number of forecast steps and 'alpha' for the confidence level....
Read full answer See technical action

The `percentile.assess` action is used to assess and compare models. It provides functionalities for calculating percentiles, box plot values, and generating various model assessment statistics....
Read full answer See technical action

To run the `assess` action, you must specify the `table` parameter, which defines the input data table, and the `response` parameter, which indicates the response variable for the model assessment....
Read full answer See technical action

You can specify the event of interest using the `event` parameter, which takes the formatted value of the response variable that represents the event. If you don't specify this and the response is numeric, the action performs an assessment for a regression model....
Read full answer See technical action

Yes. To get ROC (Receiver Operating Characteristic) data, set `includeRoc` to TRUE and specify an output table with the `rocOut` parameter. For Lift data, set `includeLift` to TRUE and use the `casOut` parameter to specify the output table. You can control the granularity of ROC calculations with `c...
Read full answer See technical action

The `assess` action supports two algorithms for percentile analysis, specified via the `method` parameter: 'ITERATIVE' (the default) and 'EXACT'. The iterative method's convergence can be fine-tuned using the `maxIters` and `epsilon` parameters....
Read full answer See technical action

You can save the fit statistics to a separate CAS output table by using the `fitStatOut` parameter. This is particularly useful when assessing nominal response variables where you must also specify the probability events (`pEvent`) and probability variables (`pVar`)....
Read full answer See technical action

The assessBias action calculates bias metrics for predictive models....
Read full answer See technical action

The 'code' parameter specifies the DATA step code that describes the model or specifies the DS2 code that is used along with an analytic store that you specify in the modelTable or modelTables parameter....
Read full answer See technical action

The 'cutoff' parameter specifies the cutoff for the confusion matrix. The default value is 0.5 and the range is (0, 1)....
Read full answer See technical action

The 'event' parameter specifies the formatted value of the response (target) variable that represents the event of interest....
Read full answer See technical action

You can use the 'frequency' parameter to specify the variable that contains frequency values....
Read full answer See technical action

The 'modelTable' parameter specifies the input table that contains the model to explain. This table must contain an analytic store or DATA step scoring code....
Read full answer See technical action

The 'modelTables' parameter is used to specify the input tables that contain the model to explain when the model is composed of analytic stores and requires accompanying DS2 code specified in the 'code' parameter....
Read full answer See technical action

The 'modelTableType' parameter specifies the type of scoring the model table contains. It can be 'ASTORE', 'DATASTEP', or 'NONE'. The default is 'ASTORE'....
Read full answer See technical action

The 'nBins' parameter specifies the number of bins to use in lift calculations. The default is 20, and the range is from 2 to 100....
Read full answer See technical action

Use the 'predictedVariables' parameter, which is a required list of variables that contain the model's predictions. The order of variables must match the order in the 'responseLevels' parameter....
Read full answer See technical action

The 'referenceLevel' parameter specifies the reference level for the sensitive variable, which is used as a baseline for bias calculations....
Read full answer See technical action

The 'response' parameter is used to specify the response (target) variable for the model....
Read full answer See technical action

The 'responseLevels' parameter specifies the list of formatted values of the response (target) variable. The order must match the 'predictedVariables' parameter. If 'event' is not specified, the first value is assumed to be the event....
Read full answer See technical action

The 'rocStep' parameter specifies the step size to use for the receiver operating characteristic (ROC) calculations. The default is 0.05....
Read full answer See technical action

The 'scoredTable' parameter specifies the output table to contain the scored outputs....
Read full answer See technical action

The 'selectionDepth' parameter specifies the depth to use in lift calculations. The default value is 10....
Read full answer See technical action

The 'sensitiveVariable' parameter is required to specify the sensitive variable to use in bias calculations....
Read full answer See technical action

The 'table' parameter is required and is used to specify the input data table for the action....
Read full answer See technical action

The assumeRole action allows a user to assume a specific administrative role, such as 'ACTION', 'DATA', or 'SUPERUSER'....
Read full answer See technical action

The available roles are 'ACTION' for unrestricted access to actions, 'DATA' for unrestricted access to data definitions, and 'SUPERUSER' which combines the privileges of both 'ACTION' and 'DATA' roles, plus the ability to manage roles and paths....
Read full answer See technical action

The 'SUPERUSER' role provides the combined privileges of the 'ACTION' and 'DATA' roles, along with the ability to manage other roles and paths....
Read full answer See technical action

The default value for the 'adminRole' parameter is 'SUPERUSER'....
Read full answer See technical action

The 'ACTION' role provides unrestricted access to action sets and actions, while the 'DATA' role provides unrestricted access to caslib, table, and column definitions, including the ability to add caslibs....
Read full answer See technical action

The `attribute` action is used to manage extended table attributes. This allows for adding, updating, dropping, converting, or exporting metadata associated with a table or its columns....
Read full answer See technical action

The `task` parameter allows you to specify one of the following tasks: 'ADD' to add new attributes, 'CONVERT' to convert attributes, 'DROP' to remove attributes, 'EXPORT' to export attributes to an XML format, and 'UPDATE' to modify existing attributes. The default task is 'ADD'....
Read full answer See technical action

You can specify attributes in three ways: 1) Using the `attributes` parameter, which takes a list of key-value pairs. 2) Using the `xml` parameter, which accepts an XML document as a string. 3) Using the `xmlPath` parameter, which specifies the path to a file containing the attributes in an XML form...
Read full answer See technical action

The `name` parameter is required and specifies the name of the in-memory table to which the attributes apply. The `table` parameter (aliased as `attrTable`) specifies the name of an existing table that stores the extended attributes, which is used for tasks like ADD, UPDATE, or CONVERT....
Read full answer See technical action
Copyright © WeAreCAS 2023-2025