I am assuming python3.5.
Official. You can do as follows. (90MB)
docker run -it python:3.5-alpine
--For 2.7, it's in testing (https://pkgs.alpinelinux.org/packages?name=py-numpy&repo=testing&arch=all&maintainer=all). --Even in 3.5, you can do your best based on here. --There is a better way. You can use miniconda by using frolvlad / alpine-glibc that enables glibc with alpine.
If you make an image with the Dockerfile below, it is 136MB, but you can easily install both numpy and scipy with "conda install numpy scipy".
Dockerfile
FROM frolvlad/alpine-glibc
ENV PATH=/opt/conda/bin:$PATH \
LANG=C.UTF-8 \
MINICONDA=Miniconda3-latest-Linux-x86_64.sh
RUN apk add --no-cache bash wget && \
wget -q --no-check-certificate https://repo.continuum.io/miniconda/$MINICONDA && \
bash /Miniconda3-latest-Linux-x86_64.sh -b -p /opt/conda && \
ln -s /opt/conda/bin/* /usr/local/bin/ && \
rm -rf /root/.[acpw]* /$MINICONDA /opt/conda/pkgs/*
CMD ["bash"]
I made tsutomu7 / alpine-python for scientific computing. (782MB)
For Kitematic, "tsutomu7 / alpine-python: jupyter" is convenient to use with a click.
that's all
Recommended Posts