Python环境配置中的报错及解决方案

(1) centos下使用yum 安装pip失败
  

   

    http://www.cnblogs.com/Rsean/articles/5635904.html
   
   

    

     安装pip

    
    

     
     

      
yum install -y python-pip
     
    
   
   

    (2) 安装过程中“
    RuntimeError: Broken tooolchain:cannot link a simple c program"
   
   

    
   
   

   

    
   
   

   

    

     
安装gcc：yum install gcc 
安装g++：yum install gcc-c++
    
    

     
    
    

     (3) 报错“ SystemError: Cannot compile 'Python.h'. Perhaps you need to install python-dev|python-devel.”
    
    
yum install python-devel
    

   
   

    
   
   

    (4) 若用到pyspark ,报错：
   
   

    

      “File "DT.py", line 6, in <module>
    
    

         from pyspark import SparkContext
    
    

     ImportError: No module named pyspark
    
    

     

      ImportError: No module named py4j.java_gateway“
     
     

      实际上，spark包里面带有pyspark , importError是由于环境变量配置问题
     
     

      
环境变量设置
vi /etc/profile
export JAVA_HOME=/usr/local/jvm/jdk1.7.0_71
export HADOOP_HOME=/home/zkpk/hadoop-2.6.5
export SPARK_HOME=/home/zkpk/spark-2.0.1-bin-hadoop2.6
export HIVE_HOME=/home/zkpk/apache-hive-0.13.1-bin
export MAHOUT_HOME=/home/zkpk/apache-mahout-distribution-0.10.1
export PYTHONPATH=$SPARK_HOME/python:$SPARK_HOME/python/lib/py4j-0.10.3-src.zip:$PYTHONPATH
export PATH=$PATH:$JAVA_HOME/bin:$JAVA_HOME/jre/bin:$HADOOP_HOME/bin:$SPARK_HOME/bin:$HIVE_HOME/bin:$MAHOUT_HOME/bin
export CLASSPATH=.:$JAVA_HOME/lib/tools/jar:$JAVA_HOME/dt.jar
#生效配置文件
source /etc/profile
      
(5)使用到bson, 报错“
      I
      mportError: No module named bson”
     
     

      问题：bson与pymongo冲突： 
      http://stackoverflow.com/questions/35254975/import-error-no-module-named-bson
     
    
   
  

sudo pip uninstall bson
sudo pip uninstall pymongo
sudo pip install pymongo

Solution 1:As stated by Martin Pieters - use the backport.

Add counter.py at /lib64/python2.6/ - this is where the collections.py is ./lib64/python2.6/collections.py Patch collections.py with:

from counter importCounter

Solution 2:use the backport_collections package. Next patch (the import statement) the package you're getting exception at i.e. pandas in my case:

from backport_collections importCounter

Solution 3: 升级python2.6 到python 2.7

  

   

    

     http://www.centoscn.com/image-text/config/2015/0910/6152.html
    
   
   

    升级遇到的问题：
   
   

    1. 如下命令更改软连接后，python 找不到
   
   

    
ln -s /usr/local/python2.7 /usr/bin/python
    vim /etc/profile 增加python2.7的环境变量
   
   2. 安装setuptools（区分版本）
   

    报错：
    python内置的urllib模块不支持https协议的解决办法
   
   

    

     http://blog.csdn.net/zyz511919766/article/details/25049365
    
    

     安装
      openssl-devel，并重新编译python
    
    

     sudo pip numpy install 仍然报错
    
    

     3. 更改软链接的方式或多或少会对系统有一定影响，建议采取下面方法
    
   
   

    https://segmentfault.com/a/1190000000654227
   
  
  

   （1）下载Python2.7.13的包
  
  

   
tar -zvxf Python-2.7.13.tgz
cd Python-2.7.13
./configure --prefix=/usr/local/python2.7
make && make install
  
  

   

  
(2)安装 setuptools + pip
  

   
# First get the setup script for Setuptools:
wget --no-check-certificate https://bootstrap.pypa.io/ez_setup.py
# Then install it for Python 2.7 :(可能会遇到usrlib问题，见2.解决方案）
cd /usr/local/python2.7/bin
./python2.7 ez_setup.py
# Now install pip using the newly installed setuptools:
./easy_install-2.7 pip

# With pip installed you can now do things like this:
pip2.7 install [packagename]
pip2.7 install --upgrade [packagename]
pip2.7 uninstall [packagename]
  
（3）使用 virtualenv
  

   
# Install virtualenv for Python 2.7 and create a sandbox called my27project:
pip2.7 install virtualenv
virtualenv-2.7 my27project

# Check the system Python interpreter version:
python --version
# This will show Python 2.6.6

# Activate the my27project sandbox and check the version of the default Python interpreter in it:
source my27project/bin/activate
python --version
# This will show Python 2.7.X
deactivate
python虚拟环境的好处：http://liuzhijun.iteye.com/blog/1872241
（4） 在PYCHARM中添加环境File -> Settings -> Project Settings -> Python Intercepter -> Python Intercepters 在 Paths 选项卡的列表中添加自定义的导入路径即可。