Description

===== pypet ===== The new python parameter exploration toolbox. pypet manages exploration of the parameter space and data storage for you. Currently supported storage is HDF5! ------------------------ What is pypet all about? ------------------------ Whenever you do numerical simulations in science, you come across two major problems. First, you need some way to save your data. Secondly, you extensively explore the parameter space. In order to accomplish both you write some hacky IO functionality to get it done the quick and dirty way. Storing stuff into text files, as MATLAB m-files, or whatever comes in handy. After a while and many simulations later, you want to look back at some of your very first results. But because of unforeseen circumstances, you changed lots of your code. As a consequence, you can no longer use your old data, but you need to write a hacky converter to format your previous results to your new needs. The more complexity you add to your simulations, the worse it gets, and you spend way too much time handling your data and results than doing science. Indeed, this was a situation I was confronted with pretty quickly during my PhD. So this project was born. I wanted to tackle the IO problems more generally and produce code that was not specific to my current simulations, but I could also use for future scientific projects right out of the box. The python parameter exploration toolkit (*pypet*) provides a framework to define *parameters* that you need to run your simulations. You can actively explore these by following a *trajectory* through the space spanned by the parameters. And finally, you can get your results together and store everything appropriately to disk. Currently the storage method of choice is HDF5. Cheers, Robert