[feature] Random sampler

[b-barton]

Is your feature request related to a problem?

We want to add or remove profiles randomly.

Describe the solution you'd like...

Modify the sampler to give it functionality to add a given number of random profiles when no profile dataset is provided or removes a percentage of profiles when provided with a profile dataset.
The random distribution could be fully random or specified with the ARGO distribution.

Describe any alternatives you've considered...

No response

Additional context

No response

[b-barton]

Should this make a random set of profile locations (from probability distribution) then sample model grid in normal way or take random sample of model grid points (from probability distribution)? Do I modify sampler_nearest_neighbour.py or make a second sampler.

[b-barton]

I decided it was sufficiently different to make a different sampler class.