Blog

  • Python Read Data in Parallel

    If the original data is spread out in a number of files, we can read them in parallel (on multi-core) machines to save time. Ideally. However, it seems a bit tricky with python.

  • How to run Jupyter Notebook on a server

    We can run Jupyter Notebook remotely on a server to run large computing jobs, which cost a lot of memory, CPU or time. You can log in the remote server, e.g.,

  • Parallel Computing using Swift and Metal

    In 4 weeks of part-time effort, with zero knowledge in Swift/Metal but the help of the Internet and curiosity, I evolved a sample code found on the Internet into a full feature and easy-to-use/modify code for one of my research problems requiring large scale parallel computing. Please feel free to check out the Swift/Metal code from my github repository and let me know any feedback you may have.

  • How to use pgfplots table in Latex

    A few commonly used methods to construct tables in latex.

  • How to use pgfplots in Latex

    This article introduces a few commonly used methods for plotting in latex. In this way, there is no need to save picture and include it. Every time when you change your numerical experiment, to update the plot, just need to update the data file.

  • How to run Java program on a server

    We aim to execute a Java Program for simulation using SSJ on server. It will be particularly helpful if the simulation takes a long time to run. This program needs two external libraries ssj.jar and optimization-1.1.jar. We do this job in the following 3 steps.

    • Step 1. Copy files from local to server
    • Step 2. Execute Java Program on server
    • Setp 3. Copy output files from server to local
  • Use git for version control

    The following is a very brief introduction. Codecademy just launched a new course on Git, the industry standard for version control. Following the link to learn in on Codecademy

  • Install Python on Mac OS X

    Installing Python and its packages on Mac is neat, thanks to Brew and pip. There are alternative ways, e.g., using the graphic installation packages provided by Enthought or Anaconda. But they mess up the system in several ways (e.g., Enthough messed up the system path which could affect brew).

  • Latex coding style

    Often in time, I find many latex source files in a mess, making editing an excruciating experience. This is not necessarily the case if we can tide up the source file. Here are some coding style I suggest, that will make the source code neat, readable and not so painful to do editing.

  • Emacs for Latex on Mac OS X

    Vim users, please cool down :) I am not saying Emacs is any better than vim. Just for the task of latex editing, Emacs combined with auctex indeed provides great convenience, at least in my view.