Configuring deep learning / development environment for machine learning in Ubuntu18.04
1, download and install Anaconda
Download: https: //www.anaconda.com/distribution/#linux
installation steps:
1) Open the terminal to download path anaconda Run:
bash
~
/Downloads/Anaconda3-5
.2.0-Linux-x86_64.sh
2) Record path configuration during installation
Prefix=/home/your name/anaconda2/
Where "your name" is your username
3) After installation setting environment variables, execute the command in the terminal:
sudo gedit ~/.bashrc
In the open .bashrc in after performing the above command with the following statement
# added by Anaconda3 installer
export PATH="/home/shi/anaconda3/bin:$PATH"
Then enter the command in the terminal, enable the configuration:
source ~/.bashrc
4) Restart ubuntu system, open a terminal, the input command
python
After entering the above command appears on the terminal of FIG indicating successful installation Anaconda
figure 1:
2, possible problems:
1) Use conda install *** required to install the package (*** as a package), the output terminal Conda command appears after installation at / home / your name / anconda3 / bin / directory module is not Conda present: 2:
figure 2:
Solution: Reinstall anaconda.
Cons: You can not import the package they need, once introduced into the package they need, but also the problems mentioned above.
2) is not introduced until the desired input command packet in its own terminal: spyder, open IDEA, the error will appear as shown in Figure 3:
image 3:
Error name: segmentation fault (core dumped)
Solution: After installing the anaconda enter the following command in the terminal, all modules update conda
conda update --all
If not, then execute the command:
sudo apt-get update
sudo apt-get upgrade
After completion of the above command to open spyder and other modules in the terminal, the error can not find this module will appear (/ home / your name / presence of these modules anconda3 bin / the / directory)
Solution: Terminal enter the command:
conda install pyopengl
At this point, the terminal input spyder, to open spyder.
Summary: 1) installation anconda
2) Update all modules conda
3) Download module pyopengl
4) Open spyder