Input Script File

Input Script File#

Non-Interactive Mode with Script Files (Optional)

When you run OpenSees or OpenSeesPy with a script file as the first argument, the program will exit automatically once the script finishes executing. This is known as Non-Interactive Mode — the file is read from top to bottom, and all commands are executed without any additional user input.

This is the recommended and most common method for running OpenSees, especially for batch jobs and large analyses.

Even on interactive platforms like Jupyter, if you run an entire script in one command, you’re still operating in non-interactive mode. However, you can combine the convenience of an input file with the flexibility of an interactive workflow by loading OpenSeesPy commands directly into notebook cells. This approach lets you develop and refine your models while integrating graphics, plots, and visualizations alongside your code.

Note: For large-scale runs on HPC systems, non-interactive mode is the default and most reliable execution style.

Use Cases#

  • Full model execution

  • Reproducible simulations

  • Running jobs on HPC or cloud environments

Sequential Execution#

Language

Command

Tcl

OpenSees inputfile.tcl

OpenSeesPy

python inputfile.py (import OpenSeesPy inside inputfile.py)

Parallel Execution (N cores)#

Use mpiexec, mpirun, or ibrun to launch distributed memory jobs:

Language

Command

Tcl (MP)

mpiexec -np N OpenSeesMP inputfileMP.tcl

Tcl (SP)

mpiexec -np N OpenSeesSP inputfileSP.tcl

OpenSeesPy (MPI)

mpiexec -np N python inputfile.py *

  • (import OpenSeesPy inside. And script must use mpi4py or similar)

  • N is the number of concurrent processors

  • Use mpiexec in JupyterHub

  • Use ibrun in the HPC system, such as Stampede3

Contents