Example Optimisation Run

This directory holds an example setup. The script model is used to either generate synthetic data to be fitted or compute the misfit or residuals for a given set of parameters. In this case the model is a quadratic surface:

\[f(x,y) = ax^2+by^2+cxy+dx+ey+f\]

When the script is used to generate a synthetic data set it generates a random parameter set (uniformly distributed between minimum and maximum parameter value) that is stored in another file. Noise is added to the synthetic data before storing it in a file.

There are two options for running the optimisation:

1. [nlopt](https://nlopt.readthedocs.io/en/latest/) which expects an objective function that returns a single value. The optimisation is configured in the example-nlopt.cfg configuration file and run it using the example.sh nlopt shell script.

2. [dfols](https://github.com/numericalalgorithmsgroup/dfols) which expects an objective function that returns a vector containing residuals. It is configured in the example-dfols.cfg configuration file and run with the `example.sh dfols</example/example.sh`_ shell script.

The install.sh shell script creates cylc workflows for each optimser.