I am passionate about finding the best tool for the job. I can get quite picky. In fact, I can get so picky that I spend a significant time researching the best tool for the job instead of working on the problem at hand. In the end, I become also passionate about the tool since I put so much thought and time into my selection process that I become sort of a marketer for it. And the time invested in researching the right tool pays off handsomely in the long-run.
The following are my tools of choice for my current work needs. This list will not include packages that address specific mathematical problems. I stay one level above that.
Operating SystemText editor: Emacs- IDE: Emacs
- Languages:
- Python for everything
- numba for targeted compilation
- R to access statistical functionality
- C for performance
- CFFI for Python-C interfacing
- IPython for interactive sessions
- Management:
- source: Git
- file sharing: BitTorrent Sync
- Viz:
- matplotlib
- whatever is on R
- Data Stores:
- HDF5
- sqlite3
- InfiniDB
- Symbolics:
- mathematica
- sympy
- Operating System:
Oh and I've got a mobile workstation solution:
- Laptop: Vizio CN15-A* An amazing value. For about a grand usd you get a 1080 IPS screen, a GPU, and a largish screen and no crapware.
- Laptop Cooler: Targus Chill Mat with 4-port USB
Logitech MK605 Notebook kit for ergonomics - Monitor: Asus MB168B+ A portable 1080 16" USB monitor. I was impressed with the performance considering that the video was going over USB.
- Monitor: Asus PA248 Gives me some more horizontal pixels than "HD" rez. (next monitor will be 4K)
Check it out:
This comment has been removed by the author.
ReplyDeleteHi Majid, your research on RNN detecting anomolies is very interesting, would you mind sharing the code of the project? thanks a bunch!
ReplyDeletereference "Recurrent Neural Networks Can Detect Anomalies in Time Series"