DockerPythonSciMac

Docker python with machine learning libs on Mac OS

In dockerfile

Python 3
libs: numpy scipy matplotlib scikit-learn pandas

Construct image

$ git clone https://github.com/CrystalRanita/DockerPythonSciMac
$ cd DockerPythonSciMac
$ docker build -t pythonml:v1 .
(where pythonml is repository name and v1 is tag)

Check docker images:

$ docker images find which one you want to run

Run container:

docker run -v /Users/host_crystal/:/home/container_crystal -it pythonml:v1 /bin/bash (where pythonml is repository name and v1 is tag.
Use -v here mount host location to container's location, modify will kept . and can easy use visual studio code to edit code files.)

Check libs version in container

After container running

python

import numpy numpy.version.version '1.15.1'

import scipy scipy.version.version '1.1.0'

import matplotlib print('matplotlib: {}'.format(matplotlib.version)) matplotlib: 2.2.3

import sklearn print('sklearn: {}'.format(sklearn.version)) sklearn: 0.19.2

import pandas print('pandas: {}'.format(pandas.version)) pandas: 0.23.4

Copy host file to container

Instruction: docker cp SourceDir containerID/DestDir
Example: docker cp MachineLearning/. 069f40db12a2:/target
Copy host file in MachineLearning directory to container's /target directory

Commit change to keep data

if use $ docker run -t -i pythonml:v1 /bin/bash .
without mount point, after exit container modify will gone without commit. Instruction: docker commit containerID IMAGE
docker commit ce7f5ced748c pythonml:v1

CrystalRanita / DockerPythonSci