Altair HyperStudy™ 2021.0 Feature Review
This article contains main features and improvements in 2021.0 release:
-
Get extensible sampling and best fits automatically
A new approach, Sampling Fit, enables users to get both appropriate sampling and best fits within a single process. By combining space-filling DOE and Fit Automatically Selected by Training (FAST) method, it automatically samples the design space and builds a fit model until desired cross-validation R-squared value is achieved.
-
Analyze the results in a more straightforward way
In Post-Processing, a new capability Side-by-Side, allows to simultaneously visualize results of different forms within the same tab. User can choose window layout and manages selections for each active window individually.
-
Leverage in automatic connection to Altair Inspire Studio™
In Define Models, a new Inspire Studio model type is now available. It allows to perform geometric changes directly within HyperStudy.
-
Control multiple models’ execution
In Define Models, user is now allowed to control the sequence of model’s execution by defining conditions such as: File Exists; Run Before; Run After; Ignore Dependencies. This capability can be particularly useful in the case of Multiphysics optimization studies for instance to avoid running time-consuming models if at least one of the analysis has failed.
-
Get automatic preview in Expression Builder
In Define Output Responses, Expression Builder and Data Source Builder dialogs have been redesigned to accommodate both the expression and its preview. Preview update is automatic to provide more intuitive guidance when setting up an expression.
-
Display linked variables in Trade-Off Panel
Linked variables can now be displayed in Trade-Off panel upon request.
-
Get an alternate view in Pareto Plot
An alternate view has been added in Pareto Plot to help user in getting better understanding of positive/negative effects.
-
Maximize levels for discrete variables in DOE
One-click option has been added to maximize levels for discrete variables in applicable screening DOE methods. Maximum and minimum levels along with values are displayed in table form.
-
Save time in evaluating independent models
In Evaluate, a new option Multi-model Parallelism, allows running simultaneously multiple independent models. An extra time saving can be obtained by combing both Multi-execution and Multi-model Parallelism.
And many more:
Enhancements in AREA Tool
New Layout for Solver Input Argument Dialog
Monitor HyperStudy in Batch Mode During Evaluation
Improved Computational Efficiency of HyperStudy Fit Model
Periodic Report Generator
Registering Python Function via GUI
Happy Exploring with Altair HyperStudy™ 2021 !