csvpandas - A wrapper of the Pandas high performance data analysis library to view and manipulate csv files

• Chris Rosenthal

This is strictly an experimental package and potentially full of bugs. I am still in the process of planning and writing up specifications. The motivation is to take advantage of Python Pandas high performance libraries for manipulating csv files on a file system.

• Python 2.7.x
• setuptools
• Pandas 0.16.2

To install csvpandas and python dependencies, run setup.py or pip from the project directory:

% cd csvpandas
% python setup.py install
# or
% pip install -U .

If you don't want to install the dependencies (numpy and pandas take a while to compile), use:

% pip install --no-deps -U .

Numpy and pandas require many dependencies to compile (and you'll likely need to compile them because versions in package managers are typically out of date). Fortunately, these can pretty easily be installed on Ubuntu 12.04 by running:

% sudo apt-get build-dep python-numpy python-pandas

Unit tests are implemented using the unittest module in the Python standard library. The tests subdirectory is itself a Python package that imports the local version (ie, the version in the project directory, not the version installed to the system) of the package. All unit tests can be run like this:

% ./testall
...........
----------------------------------------------------------------------
Ran 11 tests in 0.059s

OK

A single unit test can be run by referring to a specific module, class, or method within the tests package using dot notation:

% ./testone -v tests.test_utils

To build the Sphinx docs:

(cd docs && make html)

And to publish to GitHub pages:

ghp-import -p docs/_build/html

(ghp-import and Sphinx are both included in the requirements.txt)

Copyright (c) 2015 Chris Rosenthal

Released under the GPLv3 License