# Setup

FAQs regarding setup of an optimization problem in LS-OPT

- Strategies for metamodel-based optimization
- There are three strategies for automating the metamodel-based optimization procedure. These strategies apply to the tasks Metamodel-based Optimization and RBDO, and are available in the Task dialog.
- Strategy Single Iteration vs. Sequential
- The two strategies for the calculation of global metamodels Single Iteration and Sequential are investigated.
- Which meta model to choose?
- Remarks and hints on selection of Meta-Model
- Constrained space filling - Reasonable Design Space
- In this part, we’ll look deeper into how the space filling point sampling algorithm works. Also, the definition of sampling constraints will be explained.
- Can LS-OPT handle discrete variables?
- LS-OPT can handle discrete variables and even discrete variables in combination with continuous variables, as well as discrete variables with string values.
- How can I ensure that LS-OPT replaces the variables in the input file by integer values?
- I defined a discrete variable with a list of integer values, but LS-OPT replaces the variables in the input file by e.g. 1. with a dot.
- How can the number of solver jobs to be run in parallel be defined?
- Specifying Computing Resources for Concurrent Processing
- Is it possible to skip the reading of large include files?
- The reading of big files can be skipped in the following way:
- How can I determine the time value where a history curve takes a specific value?
- Using Lookup, LookupMin and LookupMax
- Result extraction for LS-DYNA input files using *CASE
- How can I define in LS-OPT which output files are used for result extraction if I have mutliple output result files because of *CASE?
- Probabilistic Variables: What is the difference between control and noise variables?
- A probabilistic variable is completely described using a statistical distribution. The statistical distribution defines the mean or nominal value as well as the variation around this nominal value.