DockerfileJupyter

#
#    LEAN Jupyter Docker Container 20190206
#

# Use base system for cleaning up wayward processes
FROM quantconnect/lean:foundation

MAINTAINER QuantConnect <contact@quantconnect.com>

# Install Tini
RUN wget --quiet https://github.com/krallin/tini/releases/download/v0.10.0/tini && \
    echo "1361527f39190a7338a0b434bd8c88ff7233ce7b9a4876f3315c22fce7eca1b0 *tini" | sha256sum -c - && \
    mv tini /usr/local/bin/tini && \
    chmod +x /usr/local/bin/tini

# Install Lean/PythonToolbox
RUN git clone https://github.com/QuantConnect/Lean.git && cd Lean/PythonToolbox && \
    python setup.py install && cd ../.. && rm -irf Lean

#Install Jupyter and other packages
RUN conda install -c plotly plotly=4.1.0
RUN conda install -c conda-forge jupyterlab

#Install ICSharp (Jupyter C# Kernel)
RUN wget https://cdn.quantconnect.com/icsharp/ICSharp.Kernel.20180820.zip && \
    unzip ICSharp.Kernel.20180820.zip && rm -irf ICSharp.Kernel.20180820.zip && cd icsharp && \
    jupyter kernelspec install kernel-spec --name=csharp && cd ..

# Setting some environment variables
ENV WORK /root/Lean/Launcher/bin/Debug/
ENV PYTHONPATH=${WORK}:${PYTHONPATH}

# Copy Lean files to convenient locations
COPY ./Launcher/bin/Debug/* ${WORK}

RUN find ${WORK} -type f -not -name '*.dll' -not -name '*.ipynb' -not -name '*.csx' -not -name 'decimal.py' -delete

# Copy pythonnet binaries for jupyter
COPY ./Launcher/bin/Debug/jupyter/* ${WORK}

# Setup config file
RUN echo "{ \"data-folder\": \"/home/Data/\", \"composer-dll-directory\": \"$WORK\", \"algorithm-language\": \"Python\", \"messaging-handler\": \"QuantConnect.Messaging.Messaging\", \"job-queue-handler\": \"QuantConnect.Queues.JobQueue\", \"api-handler\": \"QuantConnect.Api.Api\" }" > ${WORK}config.json

EXPOSE 8888
WORKDIR $WORK

ENTRYPOINT [ "/usr/local/bin/tini", "--" ]
CMD jupyter lab --ip='0.0.0.0' --port=8888 --no-browser --allow-root

# List packages
RUN conda list

# Usage:
# docker build -t quantconnect/lean:foundation -f DockerfileLeanFoundation .
# docker build -t quantconnect/jupyter -f DockerfileJupyter .
# docker run -it --rm -p 8888:8888 -v (absolute to your data folder):/home/Data:ro quantconnect/jupyter