Kyso's default docker image

eoin profile picEoin Murray

Here are details on the Kyso docker image that we run when you start a new container on Kyso. Its a comprehensive image for doing data-science and machine learning with python.

With Kyso you can pick and choose as many CPU's and GPU's as you want, and we give you a ready-made environment for doing data-science with python (python3). We include many typical programs like git and nano, and many python libraries like Jupyter, Tensorflow, Numpy, Scipy and Pandas. We also include plotting libraries like Matplotlib, Seaborn, Plotly and Bokeh.

Here is a list of relevant programs and python libraries installed on the environment. Contact us on laura@kyso.io if you want to request extra programs/libraries.

python-pip
python
python-dev
python3
python3-dev
python3-pip
git
gfortran
libblas-dev
liblapack-dev
libatlas-base-dev
libxml2-dev
libxslt1-dev
libreadline6
libreadline6-dev
build-essential
curl
libjpeg8
libjpeg8-dev
libfreetype6
libfreetype6-dev
zlib1g-dev
pkg-config
libffi-dev
python-pip
curl
wget
gcc
libhdf5-serial-dev
nano

Here is a list of all the python libraries installed, currently they all get installed using the 'pip3' command, so will work best with python3.

numpy
jupyter
pyzmq
scipy
pandas
pandas-datareader
matplotlib
statsmodels
scikit-learn
seaborn
nltk
gensim
Flask
django
sympy
bokeh
networkx
theano
requests
beautifulsoup4
jupyter_contrib_nbextensions
jupyter_nbextensions_configurator
jupyter_kernel_gateway
ipyleaflet
bqplot
ipyvolume
pythreejs
pymc3
jupyter_dashboards
dumb-init
dash>=0.17.7
dash-renderer>=0.7.3
dash-html-components>=0.6.2
dash-core-components>=0.5.3
plotly>=2.0.12
ipywidgets===5.2.3
tensorflow
h5py
keras