Keji Li's place

SQLAlchemy, A short tutorial

Some obscure points to note when using sqlalchemy:

  1. if you want to use numpy or your engine is sqlite, use Float columns instead of Numeric columns
  2. you can change your database defining script and have it run as long as the changes don't affect the tables, i.e.: change column type between numeric and float, add relationships on existing foreign keys.
  3. sqlalchemy doesn't auto generate __init__. The default one doesn't take any args. The an easy way to write init is to have a fields tuple and iter through *varargs and **kwargs
  4. when using hierarchy on single tables, don't force NOT NULL on children's fields

Bibpdf, a simple reference manager

This is a paper management system that is light (and easy to change) but with some important features others lack. It is supposed to be used with bibtex only, but should work with anything that eats .bib files. The database I use is sqlite, but should work with whichever of your choice (require python interface) with minimal change of the code. The database structure is in 'paper_db.sql'. The script to interact with the database is in 'pdf.' file, rename it anyway you like and put in a folder easy to work in or a folder python can find. Change the paths at the beginning of the file to your corresponding paths. If you are in windows, change the first line to the python path. read more

A common pulvinar retinotopic organization for primate species

There just need to be a post on my recently accepted paper: Pulvinar Retinotopy of a Prosimian (bush baby) read more