The current version of Jupyter Notebook computes the document state at the browser side, this is a problem if you run long jobs in a remote notebook from a laptop. If you close the browser you lose all the output of the current running cell. I will explain how we solved this problem in our lab. This solution it also allows a "walkie-talkie" like real-time collaboration.
Usage example of cProfiler, line profiler and memory profiler to detect performance problems for data analysis algorithms.
Talk about different solutions/problems to parallelize in python
video: https://www.youtube.com/watch?v=RQc7KL-ne3A
Licensed under CC BY 4.0 Creative Commons.