kaggle-python-boilerplate

Kaggle python boilerplate to start with empty main.ipynb file

Installation

Python and Git are asumed to be installed.

Pip installation:

wget https://bootstrap.pypa.io/get-pip.py
python get-pip.py

Pip documentation

Installing virtualenv:

[sudo] pip install virtualenv
# or
[sudo] pip install https://github.com/pypa/virtualenv/tarball/develop

Virtualenv documentation

Virtualenvwrapper installation:

pip install virtualenvwrapper
# or (for Mac OS X El Capitan)
pip install virtualenvwrapper --ignore-installed six

source /usr/local/bin/virtualenvwrapper.sh
# You can add previous line directly to your ~/.bashrc or ~/.bash_profile

Virtualenvwrapper documentation

Jupyter installation:

jupyter requires python development libs and gcc to build ZMQ modules.

Centos:

sudo yum install python-devel gcc
# Note: You may want to install version specific libs (e.g., `python3-devel`)

Ubuntu:

sudo sudo apt-get install python2.7-dev build-essential

# run after installing dependencies described above
pip install jupyter

Creating virtualenv:

mkvirtualenv ds

Cloning the repo:

git clone https://github.com/Hiyorimi/kaggle-python-boilerplate
cd kaggle-python-boilerplate
bash start.sh
# if jupyter installation (as part of start.sh) failed on Mac OS X El Capitan, try this:
pip2.7 install jupyter
# and re-run start.sh

Usage

Just use bash start.sh or

cd ~/Path/To/Kaggle/Folder/kaggle-python-boilerplate/notebooks
workon ds
jupyter notebook