Have you noticed that lately everyone is switching to Python? Well, if you are not surrounded by geeks who spend most of their days doing engineering problem solving work with computers you may have not heard about this. And then you may not be interested in this post either. But if somehow you are, you should keep reading!
All this talk (it's not only economists by the way) about Python made me think if the effort would be worth it. Luckily, Boragan Aruoba and Fernandez-Villaderde came to the rescue and showed that it may not be worth the effort. They ran a classic economic problem (basically grid search and value function iteration, so lots of loops and big matrices) in many languages and basically found that:
Python does improve with compilers as well, but it is not faster than Matlab in any case. Hence, if you are looking for speed (rather than flexibility or low price), Python seems to be the wrong way. Full report here. This table summarizes their findings (time in seconds):
Follow me on twitter at @diedaruich
News and posts for an active mind.