Anaconda为每个项目创建一个新的环境，将项目所需要的modules和其它项目所需要的modules隔离开来，从而避免因为各个项目因为不同的modules版本不同而产生冲突。
Anaconda is an easy-to-install, free package manager, environment manager, Python distribution, and collection of over 150 open source packages with free community support.
current env
- os:ubuntu 14.04LTS
- python:2.7
- processor:Intel® Core™ i7-4710MQ CPU @ 2.50GHz × 8
- graphics:Intel® Haswell Mobile
- OS type:64-bit
installation
download & install
Go to the Anaconda download to download it.
Python 3.5 version
bash Anaconda3-4.2.0-Linux-x86_64.sh
Python 2.7 version
bash Anaconda2-4.2.0-Linux-x86_64.sh
Note:during the installation, you can choose the path where you want to install it. I choose to install it in /opt/python.
create a conda environment
This new created environment is used to store all the python libraries for your project.
$ conda create -n [project_name] python=[python_version]
- project_name is the project name(environment name),here we use stock.
- python_version : choose which python version for your project, like 2.7,3.5 and so on, here we choose 3.5.
note:
the new created env is located at:
~/.conda/env/
all the modules installed for the env is located in
~/.conda/envs/[env_name]/lib/python3.5/site-packages/
activate and deactivate the environment
activate:
$ source activate stock
deactivate:
(stock)$ source deactivate
install packages for env
First, we need to activate the environment(that’s stock here)
Then, use pip to install the packages(or sometimes can use conda to install)
use pip
- install
$ pip install [package_name] - uninstall
$ pip uninstall [package_name]
note: sometimes, pip may not work, there is still another way to install the moduel from the source code as the following:
$ python setup.py install
startup with IPython
make sure you’ve activate the tensorflow environment
(stock)$ ipython notebook
note: if this doesn’t work, use the following command to install ipython:
(stock)$ conda install ipython
or
(stock)$ pip install ipython