Basic Installation Flow
Thanks to Kenarius Octonotes, I basically followed his article Install Up-to-date Scientific Environment in Ubuntu 14.04 With Python 3.4. The installation flow goes like this (with some modification with Kenarius’ note):
#!/usr/bin/env bash # change to root sudo su # install python development packages and g++ apt-get install -y python3-dev g++ # install dependencies for scipy apt-get install -y libblas-dev liblapack-dev gfortran # install dependencies for matplotlib apt-get install -y libfreetype6-dev libpng-dev # install numpy apt-get install python3-numpy # exit from `sudo su` exit # ipython notebook has bug: # https://bugs.launchpad.net/ubuntu/+source/python3.4/+bug/1290847 python3 -m venv --clear --without-pip venv/ipython-notebook # activate virtual env and install pip source venv/ipython-notebook/bin/activate # (1) DO NOT switch to `easy_install pip` in the middle of the installation flow wget https://bootstrap.pypa.io/get-pip.py python3 get-pip.py rm get-pip.py # install scientific packages sudo -H pip3 install numpy sympy matplotlib scipy pandas # install ipython notebook and dependencies sudo -H pip3 install ipython pyzmq jinja2 pygments bokeh # install latest dev scikit-learn and build it # sudo -H pip3 install cython https://github.com/scikit-learn/scikit-learn/archive/master.zip # You don't have to use the latest version of scikit-learn sudo -H pip3 install cython scikit-learn # install prettyplotlib by Olga Botvinnik for beauty plots sudo -H pip3 install brewer2mpl prettyplotlib # install ipython and notebook, jupyter and patsy sudo -H pip3 install ipython[notebook] sudo -H pip3 install jupyter sudo -H pip3 install patsy # start notebook. current folder will be your workspace sudo ipython notebook # deactivate venv, complementary to `source venv/ipython-notebook/bin/activate` deactivate
However, there might be some problems.
Problem 1: invalid command ‘egg_info’
The solution is provided in stackoverflow: Can’t install via pip because of egg_info error:
Found out what was wrong. I never installed the setuptools for python, so it was missing some vital files, like the egg ones.
If you find yourself having my issue above, download this file and then in powershell or command prompt, navigate to ez_setup’s directory and execute the command
python ez_setup.pyand this will run the file for you.
But please don’t change to
easy_install pip, which is suggested in the above post, in the middle way.
If you did got some trouble from
easy_install pip, you can check stackoverflow: How do I remove packages installed with Python’s easy_install?.
Problem 2: No module named notebook.notebookapp
According to stackoverflow: ImportError: No module named notebook.notebookapp, the solution is:
For 4.0 and above You need to install the notebook app separately from https://github.com/jupyter/notebook
pip install jupyter
How to check the versions of your packages?
Problem 3: Unrecognized alias
The profile mechanism changed, so you don’t have to create profile under
~/.ipython/, as stackoverflow: IPython notebook won’t read the configuration file says:
IPython has now moved to version 4.0, which means that if you are using it, it will be reading its configuration from
~/.ipython. You have to create a new configuration file with
jupyter notebook --generate-config
and then edit the resulting
~/.jupyter/jupyter_notebook_config.pyfile according to your needs.
More installation instructions here.