Frequently Asked Questions

550 questions found.

Page 5 / 11

Yes, an attribute can be applied to a specific column. When using the `attributes` parameter, you can include the `column` parameter within an attribute's definition, providing the name of the column you want to associate the attribute with. If the `column` parameter is omitted, the attribute applie...
Read full answer See technical action

The `augmentImages` action is used to augment images by creating patches from them and applying various mutations or transformations. This is a common technique in computer vision to artificially expand a training dataset....
Read full answer See technical action

The required parameters are `table`, which specifies the input table containing the images, `casOut`, which defines the output table for the augmented images, and `augmentations`, which is a list specifying the cropping and mutation options....
Read full answer See technical action

You can create patches by defining one or more blocks within the `augmentations` parameter. You can either specify the exact coordinates and dimensions of a patch using `x`, `y`, `width`, and `height`, or use `sweepImage` set to TRUE to automatically create patches across the entire image using a sl...
Read full answer See technical action

Within each `augmentations` block, you can specify a `mutations` list. Available mutations include geometric transformations like `horizontalFlip`, `verticalFlip`, `rotateLeft`, and `rotateRight`, as well as appearance modifications such as `colorJittering`, `colorShifting`, `darken`, `lighten`, `sh...
Read full answer See technical action

Yes, you can apply a random rotation by specifying `rotateLeft` or `rotateRight` with a `type` of 'RANGE'. For example, `mutations={rotateRight={type='RANGE', value={0, 90}}}` would apply a random right rotation between 0 and 90 degrees....
Read full answer See technical action

You can use the `addColumns` parameter with the value `augmentAttributes`. This will add columns to your output table that provide details about the augmentation, such as `_patch_x_`, `_patch_y_`, `_patch_width_`, and `_patch_height_`....
Read full answer See technical action

Yes, you can set the `useWholeImage` parameter to TRUE within an `augmentations` block. This will apply the specified mutations to the entire image instead of a cropped patch....
Read full answer See technical action

The bart.bartGauss action fits Bayesian additive regression trees (BART) models to normally distributed response data....
Read full answer See technical action

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

The 'model' parameter is used for this. It requires a 'depVars' subparameter for the target variable and an 'effects' subparameter to list the explanatory variables....
Read full answer See technical action

The 'nTree' parameter specifies the number of trees in the sum-of-trees ensemble. Its default value is 200....
Read full answer See technical action

The 'nBI' parameter sets the number of burn-in iterations (default 100) which are discarded. The 'nMC' parameter sets the number of subsequent MCMC iterations that are saved for prediction (default 1000)....
Read full answer See technical action

You can use the 'store' parameter to save the model in a binary table object. This object can then be used for scoring new data with the bartScore action....
Read full answer See technical action

The 'missing' parameter controls this behavior. The default is 'SEPARATE', which treats missing values as a distinct category. Other options include 'NONE' (exclude observations), 'MACBIG', and 'MACSMALL' (treat as largest/smallest machine value)....
Read full answer See technical action

The bart.bartProbit action is designed to fit probit Bayesian additive regression trees (BART) models specifically for binary distributed response data....
Read full answer See technical action

You can use the 'nTree' parameter to specify the number of trees in the sum-of-trees ensemble. The default value is 200....
Read full answer See technical action

The process involves a burn-in phase followed by the main sampling phase. The 'nBI' parameter sets the number of burn-in iterations (default 100), which are discarded. The 'nMC' parameter specifies the number of subsequent MCMC iterations that are saved for prediction (default 1000)....
Read full answer See technical action

The 'missing' parameter controls this. The default is 'SEPARATE', which treats missing values as a distinct group. Other options include 'NONE' (excludes observations with missing values), 'MACBIG' (treats missing as the largest machine value), and 'MACSMALL' (treats missing as the smallest machine ...
Read full answer See technical action

Yes, you can use the 'store' parameter to save the model into a binary table object. This stored model can then be used for scoring new data with the bartScore action....
Read full answer See technical action

The bartScore action creates a table on the server that contains results from scoring observations by using a fitted Bayesian additive regression trees model....
Read full answer See technical action

The `casOut` parameter is required to create a table on the server that contains observation-wise statistics computed from scoring a data table....
Read full answer See technical action

The `restore` parameter is used to specify a binary table object from a previous model fitting....
Read full answer See technical action

When `avgOnly` is set to FALSE, predictions from each MCMC sample are included in the output table in addition to the sample average predictions. By default, it is TRUE....
Read full answer See technical action

Use the `copyVars` parameter to specify a list of one or more variables to be copied from the scoring table to the output table....
Read full answer See technical action

The `alpha` parameter specifies the significance level for constructing all equal-tail credible limits, with a default value of 0.05....
Read full answer See technical action

You can use the `pred`, `resid`, `lcl` (lower credible limit), and `ucl` (upper credible limit) parameters to name the respective output statistics columns....
Read full answer See technical action

The bart.bartScoreMargin action computes predictive margins by using a fitted Bayesian additive regression trees (BART) model....
Read full answer See technical action

The 'model' parameter is required and specifies a binary table object that was created from a previous model fitting, which contains the BART model to be used for scoring....
Read full answer See technical action

You must use the 'table' parameter to specify the input data table that you want to compute predictive margins for....
Read full answer See technical action

A predictive margin is a statistical measure. In this action, you define it using the 'margins' parameter (alias: 'scenarios'). Each margin is a scenario where you can specify variables to modify and the values they should be set to, using the 'at' subparameter....
Read full answer See technical action

You can compute the difference between two predictive margins by using the 'differences' parameter (alias: 'diffs'). You need to specify the 'evtMargin' (event margin) and 'refMargin' (reference margin) by their names, which you defined in the 'margins' parameter....
Read full answer See technical action

The 'alpha' parameter specifies the significance level for constructing equal-tail credible limits for the computed margins. The default value is 0.05....
Read full answer See technical action

The `batchresults` action is used to change a currently running action to batch results mode....
Read full answer See technical action

The `batchresults` action requires the `uuid` parameter, which is a string that specifies the UUID of the session....
Read full answer See technical action

To use the `batchresults` action in CASL, you would call `session.batchresults` and provide the session's UUID. For example: `session.batchresults / uuid="your-session-uuid";`....
Read full answer See technical action

Yes, the example provided in the documentation, 'Switch an Action Running in Another Session to Batch Mode,' indicates that you can use the `batchresults` action to change the mode of an action in another session by specifying its UUID....
Read full answer See technical action

The biconnectedComponents action is used to calculate the biconnected components and articulation points of a graph within the Network Optimization action set....
Read full answer See technical action

A biconnected component is a maximal subgraph that remains connected even if any single node is removed. It's a measure of a graph's robustness....
Read full answer See technical action

An articulation point, also known as a cut vertex, is a node in a graph whose removal would increase the number of connected components, effectively splitting the graph into separate subgraphs....
Read full answer See technical action

The action generates several output tables: `out` for summary information, `outLinks` for link-level component data, `outNodes` for node-level articulation point data, and `outBCTreeLinks` and `outBCTreeNodes` for the block-cut tree structure....
Read full answer See technical action

Yes, you can use the `direction` parameter. It can be set to `UNDIRECTED` (which is the default) or `DIRECTED` to specify how the input graph should be treated....
Read full answer See technical action

The bnet action, part of the Bayesian Net Classifier Action Set, uses Bayesian network models to classify a target variable....
Read full answer See technical action

The bnet action can learn several network structures, including GENERAL (or GN), Markov Blanket (MB), NAIVE, Parent-Child (PC), and Tree-Augmented Naive (TAN) Bayesian networks....
Read full answer See technical action

For interval variables, missing values can be handled in two ways: they can be ignored, or they can be imputed with the mean of the variable. This is controlled by the 'missingInt' parameter....
Read full answer See technical action

For nominal variables, the bnet action provides three options for handling missing values: ignoring the observations, imputing with the mode of the variable, or treating the missing values as a distinct level. This is controlled by the 'missingNom' parameter....
Read full answer See technical action

The bnet action supports several methods for independence tests, specified by the 'indepTest' parameter: CHISQUARE (chi-square statistic), GSQUARE (G-square statistic), MI (normalized mutual information), CHIGSQUARE (both chi-square and G-square), and ALL (chi-square, G-square, and mutual informatio...
Read full answer See technical action

The 'maxParents' parameter specifies the maximum number of parents for each node in the network. Its default value is 5, and it can range from 1 to 16....
Read full answer See technical action

The 'saveState' parameter can be used to specify an output table where the trained model is saved for future scoring purposes....
Read full answer See technical action

The boxChart action, part of the Statistical Process Control (SPC) action set, is used to produce box charts for analyzing process variation....
Read full answer See technical action
Copyright © WeAreCAS 2023-2025