The home of the Algorist

My professional interests include earth science, numerical modelling and problem solving through code.

Jupyter notebooks are great for experimentation, reporting and sharing. In a project there are often times when you need to transition from this activity to production ready code. The easiest first step is to convert your notebook to a script. In order to do this you will need jupyter and nbconvert packages installed.

Latin hypercube sampling aims to bring the best of both worlds: the unbiased random sampling of monte carlo simulation; and the even coverage of a grid search over the decision space. It does this by ensuring values for all variables are as uncorrelated and widely varying as possible (over the range of permitted values).

This post is a draft of an article I wrote on the arcadisgen.com website…

Last week a colleague of mine shared the news that the UK Home Office has agreed to scrap its controversial ‘visa-streaming’ immigration algorithm after a successful legal challenge.