Better Analysis Workflow with [I]Python

1. Donwload and install Virtualbox and Virtualbox Extension pack which can be found on the same page. [ SLAC Mirror or OSX and Windows]
2. Download our Virtual Appliance[Catech Mirror][Google Drive Mirror](~1 GB). It's pretty big (~3-4 GB) after decompression. It's Ubuntu 12.04 with everything we need for this tutorial installed.
3. Open your virtualbox. Click [File]...[Import Appliance] and choose OVA file you just downloaded, or simple double click the OVA file from your host machine.
   • Check reinitialize MAC address or else SLAC network might get confused. If you forgot to do this, go to [Setting]...[Network]..[Advance] and regenerate the MAC address.
4. You will see pythontutorial virtual machine show up on the list. Just double click it. You will boot into Ubuntu desktop.
   • If it complains about USB driver etc., you can fix it by dowload Virtualbox Extension Pack OR disable USB by going to [Setting]...[Ports]...[USB] and uncheck Enable USB Controller.
5. The first thing you want to do after your ubuntu bootup is to change password. To change the password click terminal icon on the left hand side and type passwd. The original password is pythontutorial which is identical to the username. There is no ssh service or anything other network service running on the disk image. This is just a good security practice.
6. There is a working version of tutorial at ~/babar_python_tutorial. Open a Terminal. There is a terminal icon on the left hand side. (If you dont't find the Terminal icon, click the top-left Ubuntu "Dash Home" button and type terminal to search the app). Then follow the steps in Test it out section below (skipping git clone ..., but you may need to git pull. See that section for detail).

sudo apt-get install git subversion build-essential dpkg-dev make g++ gcc gfortran binutils libx11-dev libxpm-dev libxft-dev libxext-dev libpcre3-dev libfftw3-dev python-dev libpng-dev libblas-dev liblapack-dev t1-xfree86-nonfree ttf-xfree86-nonfree ttf-xfree86-nonfree-syriac xfonts-75dpi xfonts-100dpi

cd /usr/local/src
#or grab the latest one from their website
sudo wget ftp://root.cern.ch/root/root_v5.34.03.source.tar.gz
sudo tar xvf root_v5.34.03.source.tar.gz
cd root
sudo ./configure --prefix=/usr/local
sudo make
sudo make install

#make sure your system knows how to load root dynamic lib
sudo echo /usr/local/lib/root | sudo tee /etc/ld.so.conf.d/cern-root.conf
sudo ldconfig

sudo apt-get install python-setuptools python-pip

sudo pip install numpy
sudo pip install scipy

sudo pip install --upgrade  tornado pyzmq ipython matplotlib scikit-learn root_numpy iminuit probfit cython

1. Install Python 2.7.x (Not 3.x).
   1. You may have this install already try typing python --version. It should say something like Python 2.7.1. If you have anything older than 2.7.x then you will need to reinstall it.
   2. If you have older version, download it form here. Once you install it make sure you add /Library/Frameworks/Python.framework/Versions/2.7/bin/ to your PATH.
2. Install fortran compiler If you don't have one already.
   1. Try gfortran --version
   2. Download them 10.6, 10.7 and for 10.8
   3. brew and macports also offer their corresponding gfortran.
3. Install ROOT.
   1. You may have this installed already. Try typing root and
      root-config --incdir --libdir If you they both give you something sensible then good.
   2. If you don't have one install it.
4. Install pip.
   1. You may have it already try pip --version Make sure it says something like python 2.7 at the end.
   2. If not, easy_install --upgrade pip
5. Install sparse matrix suite
   1. Scipy requires umfpack (or other matrix routines). This will make scipy installtion script happy.
   2. Install Brew
   3. Tap homebrew-science brew tap homebrew/science
   4. Install suite-sparse brew install suite-sparse
6. Install Various Python Packages You can do all this with three simple steps.

   pip install --upgrade numpy

   pip install --upgrade scipy

   pip install --upgrade ipython  matplotlib tornado pyzmq scikit-learn root_numpy iminuit probfit

   It has been reported that scipy pip install doesn't work on OSX 10.8 follow this 10.7's instruction instead. In case you need individual package[Show]
   1. Install IPython.
      1. You may have it already try ipython --version. It should say 0.13
      2. Otherwise pip install ipython
      3. You should be able to type ipython --version
   2. Install tornado
      1. Try python -c "import tornado; print tornado.version"
      2. pip install tornado if you don't have one
   3. PyZmq
      1. Try python -c "import zmq; print zmq.__version__"
      2. pip install pyzmq if you don't have one
   4. numpy
      1. Try python -c "import numpy; print numpy.__version__"
      2. pip install numpy
   5. matplotlib
      1. Try python -c "import matplotlib; print matplotlib.__version__"
      2. pip install matplotlib
   6. scipy
      1. Try python -c "import scipy; print scipy.__version__"
      2. pip install scipy
   7. scikit-learn
      1. Try python -c "import sklearn; print sklearn.__version__"
      2. pip install scikit-learn
   8. root_numpy
      1. Try python -c "import root_numpy; print root_numpy.__version__"
      2. pip install root_numpy
   9. iminuit
      1. Try python -c "import iminuit; print iminuit.__version__"
      2. pip install iminuit
   10. probfit
       1. Try python -c "import probfit; print probfit.__version__"
       2. pip install probfit