What is this all about?
What is this all about?
The Goal of this complete project is to simulate galaxies. So the obvious two steps are: 1. Generating point clouds and 2. simulating the pointclouds that we generated.
This all started when Tim Tugendhat invited me to Heidelberg to help visualize the NFW-function. After we made som visualizations, we started generating our own clusters using the Navarro Frenk White profile. That was so much fun that I quickly started generating more and more stars and optimized the complete process.
I then started to think about stuff to do with the millions of stars I generated and realized that Simulating the clusters would be a very hard, but informative task to learn new skills.
Generating the pointclouds
In the aftermath, generating the pointclouds was one of the easier parts of the project. A detailed summary on how the stars are generated can be found in the writeup.
The process short version can be explained in the following way: Create a star, test if it should be kept or not, keep it if yes and repeat the process if not. The Process is repeated few thousand times and in the end, a clusted of objects apears on the screen. The Navarro Frenk White profile is used to determine what star is kept and what star is discarded.
Using this technique, a lot of stars can be generated, and because of all the computations running independently, this whole process scales very well. After containerizing the software generating random stars and sorting "bad ones" out, it can be thrown onto a cluster of computers who generate stars 24/7
