Project template

Projects using pkgpl_base docker images.

Requirements

• docker for cpu.
• nvidia-docker for gpu.

Docker Setting

In docker directory, add additional packages you need to Dockerfile. Set PROJECT_NAME and DEVICE in Env.sh and run following commands.

./build.sh    # docker build
./run.sh      # docker run
./exec.sh -h  # docker exec
./helper.sh   # docker images, ps, stop, rmi

Directories

Directory structure from Good enough practices in scientific computing.

• bin: put external scripts or compiled programs
• data: put raw data and metadata
• doc: put text documents associated with the project
• results: put files generated during cleanup and analysis
• src: put project source code

Directories above will be mounted in the docker container.

• path:/home/$USER/$RPOJECT_DIR/
• link:/work/$USER/$PROJECT_DIR/

Example projects

• Youtube
  • Facies classification using Machine Learning
  • Wind Data Analysis
  • Full Waveform Inversion
• Seismic Data Analysis [Youtube]