Skip to content

Calibrate a frame

The basic command to calibrate a frame to reference data in an ingest form using the optim_tool algorithm is:

1
python -m emodpy_workflow.scripts.calibrate -f FRAME -p PLATFORM optim_tool

... where FRAME is the name of the frame to calibrate and PLATFORM is the idmtools.ini platform name to run on.

There are numerous additional parameters that can be set to control the behavior of the process. They are specified before the algorithm name (like optim_tool) on the command line.

To see parameters for controlling the calibration process, run:

1
python -m emodpy_workflow.scripts.calibrate -h

Each optimization algorithm (like optim_tool) has its own parameters that are specified after the algorithm name. To see parameters for controlling the details of the chosen optimization algorithm, run:

1
python -m emodpy_workflow.scripts.calibrate optim_tool -h

Resample a calibration

Resampling a calibration is the process of selecting one or more sets of parameters from a calibration. These parameter sets are effectively "a model calibration", the end result of a calibration process. They also contain the random seed (Run_Number) to allow exact recreation of a simulation.

To resample a calibration, run the following:

1
python -m emodpy_workflow.scripts.resample -d CALIBRATION_DIR -m METHOD -n NUMBER -o FILE

... where CALIBRATION_DIR is the directory path of a calibration process that has been performed, METHOD is the sampling algorithm, NUMBER is the count of parameter sets to select, and FILE is the file path to write CSV results to.