Home
Perform Studies
A study is a self-contained project in which models, variables, output responses, and approaches are defined.
Setup the Approaches
Once the study Setup is complete, an unlimited combination of approaches can be added to a study. A study approach is a specific set of steps taken to study the mathematical model of a design.
Setup DOE Studies
A DOE is a series of tests in which purposeful changes are made to the input variables to investigate their effect upon the output responses and to get an understanding of the global behavior of a design problem. By running a DOE, you can determine which factors are most influential on an output response.
Select a Numerical Method
Select a numerical method to use when evaluating the DOE.
DOE Methods
Numerical methods available for a DOE approach.
Latin HyperCube
A square grid containing sample positions is a Latin square if, and only if, there is only one sample in each row and each column. A Latin HyperCube DOE, categorized as a space filling DOE, is the generalization of this concept to an arbitrary number of dimensions.

Welcome
What's New
View new features for HyperStudy 2024.1.
Get Started
Learn the basics and discover the workspace.
Tutorials
Discover HyperStudy functionality with interactive tutorials.
Manage Files and Data
Create, open, import, and save models.
Perform Studies
A study is a self-contained project in which models, variables, output responses, and approaches are defined.
- Setup the Study
  Before you can create approaches you must first setup your Study by defining input variables and output responses.
- Setup the Approaches
  Once the study Setup is complete, an unlimited combination of approaches can be added to a study. A study approach is a specific set of steps taken to study the mathematical model of a design.
  - Setup DOE Studies
    A DOE is a series of tests in which purposeful changes are made to the input variables to investigate their effect upon the output responses and to get an understanding of the global behavior of a design problem. By running a DOE, you can determine which factors are most influential on an output response.
    - Add a DOE Approach
      Add approach to the study.
    - Define Definition
      Define the models, input variables, and output responses to be used in the study.
    - Select a Numerical Method
      Select a numerical method to use when evaluating the DOE.
      - DOE Methods
        Numerical methods available for a DOE approach.
        Box Behnken
        Generates higher order response surfaces using fewer required runs than a normal factorial.
        Central Composite Design (CCD)
        Central Composite Design contains an imbedded factorial or fractional factorial design with center points that are augmented with a group of star points that allow the estimation of curvature.
        D-Optimal
        Primarily intended to be used as the input matrix for a Least Squares Regression Fit. By identifying the type of regression that will be used, samples are selected to maximize the determinant of the information matrix that is inverted during the Fit’s regression analysis, which in turn improves the numerical efficiency of the DOE.
        Fractional Factorial
        A factorial experiment in which only a chosen fraction of the combinations required for the Full Factorial DOE is run.
        Full Factorial
        Evaluates all possible combinations of input variable levels. This will resolve all the effects and interactions.
        Hammersley
        Hammersley sampling belongs to the category of quasi-Monte Carlo methods. This technique uses a quasi-random number generator, based on the Hammersley points, to uniformly sample a unit hypercube.
        Latin HyperCube
        A square grid containing sample positions is a Latin square if, and only if, there is only one sample in each row and each column. A Latin HyperCube DOE, categorized as a space filling DOE, is the generalization of this concept to an arbitrary number of dimensions.
        Modified Extensible Lattice Sequence (Mels)
        A lattice sequence is a quasi-random sequence, or low discrepancy sequence, designed to equally spread out points in a space by minimizing clumps and empty spaces.
        Sobol Sequence
        Sobol Sequence is a widely known low-discrepancy sequence of quasi-random sampling family. It is designed to equally spread-out points in a space by minimizing clumps and empty spaces.
        Plackett Burman (PB)
        Screens the maximum number of main effects with the least number of experimental runs in case of two-level factors.
        Run Matrix
        Load your own design matrix.
        Taguchi
        Explores how controllable variables can be used to mitigate the effects from the uncontrolled variables.
        User Defined Design
        Load your own design matrix.
      - Edit the Run Matrix
        Edit the summary of run data stored in the run matrix by editing existing runs or adding new run data.
      - Reuse Run Data
        An Inclusion matrix contains existing data that will be appended into the newly created approach as known data points. This data typically comes from other approaches, such as DOEs or previously run Optimizations.
    - Evaluate
      Run the approach.
    - Post Processing
      View the computational results from the DOE.
    - Create Reports
      Package reports for data generated during the approach.
  - Setup Fit Studies
    A Fit is a mathematical model that is trained by data and is capable of predicting output response variables for a given set of input variables.
  - Setup Optimization Studies
    An Optimization is a mathematical procedure used to determine the best design for a set of given constraints, by changing the input variables in an automatic manner.
  - Setup Sampling Fit Studies
    A Sampling Fit is a combination of space-filling DOE method and mathematical model trained by the data generated.
  - Setup Stochastic Studies
    A Stochastic approach is a method of probabilistic analysis where the input variables are defined by a probability distribution, and consequently the corresponding output responses are not a single deterministic value, but a distribution.
  - Setup Basic Studies
    A Basic approach can be used to test nominal values and bounds by performing a nominal run, system bound check, or sweep.
  - Setup Verification Studies
    A Verification approach compares two data sets in a side by side comparison.
Learn the Approaches
Each approach in HyperStudy serves a different purpose in the design study.
Customize HyperStudy
Customize HyperStudy by registering solver scripts, functions, and optimizers, and defining user preferences files.
Keyboard Shortcuts
Keyboard shortcuts used to access HyperStudy features.
Frequently Asked Questions
This section provides quick responses to typical and frequently asked questions regarding HyperStudy.

View All Altair HyperWorks Help

Latin HyperCube

A square grid containing sample positions is a Latin square if, and only if, there is only one sample in each row and each column. A Latin HyperCube DOE, categorized as a space filling DOE, is the generalization of this concept to an arbitrary number of dimensions.

When sampling a design space of N variables, the range of each variable is divided into M equally probable intervals. M sample points are then placed to satisfy the Latin HyperCube requirements. As a result, all experiments have unique levels for each input variable and the number of sample points, M, is not a function of the number of input variables.

Figure 1.

Figure 2 shows a Latin HyperCube (left) and Hammersley (right) for 100 runs.

Figure 2.

Usability Characteristics

To get a good quality fitting function, a minimum number of runs should be evaluated. (N+1)(N+2)/2 runs are needed to fit a second order polynomial, assuming that most output responses are close to a second order polynomial within the commonly used input variable ranges of -+10%. An additional number of runs equal to 10% is recommended to provide redundancy, which results in more reliable post-processing. As a result, this equation is recommend to calculate the number of runs needed or a minimum of 1.1*(N+1)(N+2)/2 runs.
The structure of a Latin HyperCube run matrix ensures that the runs are orthogonal. Orthogonality is desirable because it is less likely to result in singularities when creating Least Squares Regression fits.
Any data in the inclusion matrix is combined with the run data for post-processing. Any run matrix point which is already part of the inclusion data will not be rerun.

Settings

In the Specifications step, Settings tab, change method settings.


Parameter	Default	Range	Description
Number of Runs	$\frac{1.1 (N + 1) (N + 2)}{2}$	> 0 integer	Number of new designs to be evaluated.
Random Seed	1	Integer 0 to 10000	Controlling repeatability of runs depending on the way the sequence of random numbers is generated. 0 Random (non-repeatable). >0 Triggers a new sequence of pseudo-random numbers, repeatable if the same number is specified.
Use Inclusion Matrix	Off	Off or On	Concatenation without duplication between the inclusion and the generated run matrix.