[PYTHON] Pip the machine learning library from one end (Ubuntu)

I pip all the libraries I know. Long live pip! If you find a new library, add more and more.

Installation of pip body

$ sudo apt-get install python-pip

How to use pip

For some reason, in my environment, it doesn't work unless I use $ sudo -H pip install pip library name. Other sites etc. say $ pip install library name, so I will go here below. I wonder if there is anyone who needs sudo like me ~

Library of scientific computing

Libraries: numpy, scipy, pandas, matplotlib, scikit-image pip install numpy scipy pandas matplotlib scikit-image

Machine learning

Library: chainer, tensorflow, Theano, keras, scikit-learn pip install chainer tensorflow Theano keras scikit-learn I don't have a GPU I don't have, so I don't know CUDA or cuDNN. I want a GPU ~ Also, Caffe seems to be difficult to install. (Seen on the net)

Note only Open AI Gym

$ sudo apt-get install -y python-numpy python-dev cmake zlib1g-dev libjpeg-dev xvfb libav-tools xorg-dev python-opengl libboost-all-dev libsdl2-dev swig It seems that you can't install all of them unless you install them first. (Information from Installing everything in README.rst on Official github) And all environment installation $ sudo pip install gym[all]

Not a pip but a machine learning library

Discover the recently popular library of deep reinforcement learning keras-rl! First install git $ sudo apt-get install git After that, according to README.md (It seems better to include h5py as well) $ pip install h5py $ cd" Directory for github " $ git clone https://github.com/matthiasplappert/keras-rl.git $ cd keras-rl $ sudo python setup.py install Since the example is interesting, I will introduce it $ python" Directory for github "/keras-rl/examples/dqn_cartpole.py I'll do my best to endure it ~ There seem to be many other things. Also, press Ctrl + C to finish.

reference

Deep learning × Python I learned about keras and keras-rl by looking at this article. It doesn't seem to be very famous, but keras is nice because the code is intuitively short.

Postscript

For Python3 $ sudo apt-get install python3-pip Install pip3 with later $ pip3 install library name OK keras-rl is $ sudo python3 setup.py install OK if you change it to Personally, I'd like to change everything to Python 3 as soon as possible. The reality is that Ubuntu 16.04 LTS still has Python 2 as its standard. It's OK if you install both for the time being