Starting points

My starting point was an online course in genomic data science, which requires some coding in Python. In this blog, I’ll write about my experience of learning Python and Julia, and deep learning frameworks built for them. I’ll tell the story exactly as it happens, in raw and arduous detail. 

Mapping the Milky Way

… Is a gigantic task when you consider that the Milky Way galaxy is about 100,000 light years across, and our sun is 27,000 light years from the Galactic Center. We’re in a lonely part of the Milky Way, on the inner edge of a spiral called the Orion Arm. Yet the Milky Way galaxy is just one galaxy in a huge group of galaxies called a supercluster. A team of astronomers at the university of Hawaii have been mapping this supercluster. We now know that it’s 500 million light years in diameter and contains a hundred thousand galaxies. The Milky Way galaxy is located in the outskirts of this supercluster. Laniakea, our hone supercluster of galaxies.

Supercomputers and modeling 

The fastest supercomputers today have thousands of processors working in parallel, giving a total processing power that is measured in petaFLOPS – quadrillions of floating point operations per second. So, supercomputers are used to run complex models that require huge numbers of calculations every second. For example, supercomputers are used to run models of climate in the past. The results of running these models can then be used to make predictions about future climate changes. As well as running on supercomputers, the work of modelling climate change is also allocated to millions of personal computers around the world, via a technology known as distributed computing. The modelling task is broken down into millions of small tasks that are allocated to millions of processing units, which return the results of processing each task to the server and then receive further tasks.