2.2. Kernel / Docker Image Selection#

Here we chose a pre-defined environment, which defines available languages, software environments etc. These are often customised for a specific purpose. By default, most of them include a Command Line, and Python kernel.

_images/NewLandingPage4.jpg

Fig. 2.4 Kernel Selection Page
after first use, have to use Menu “File” ➪ “Hub Control Panel”#

Direct Linking to particular Server/Kernels

A URL parameter can be added to directly link to a “profile” (what you select later), e.g. https://jupyter.uni-bremen.de/hub/spawn?profile=fb5_datascience

Current tags (as of 2022-11-22) include:

  • generic Uni-Bremen tags:

    • standard-python, standard-r, standard-java

  • FB05 tags:

    • python, scipy, datascience, gmt, qgis

  • Tags from other FB:

    • rcourse

    • multikernel

    • testmodule

On pressing “Start” at the bottom of the selection, a server is started.

_images/kernelStartingSmall.png

Fig. 2.5 Kernel / Server Loading Indicator#

and finally presenting the main User Interface with “Launcher”:

_images/LauncherMain.png

Fig. 2.6 Kernel Selection Page
after first use, have to use Menu “File” ➪ “Hub Control Panel”#

Attention

After the initial selection of the server, another selection can be made by the menu entries File > Hub Control Panel > Stop my Server > Start my Server This will close the running kernels and allows changing to a new server option.

_images/StartStopServer.png

Fig. 2.7 Stopping and Starting a different Server.#

The launcher allows choosing between Notebook, Console or Terminal views (depending on Server selected), often for particular programming environments or languages. Special images allow usage of multiple languages in one script. A new launcher can be opened by clicking on the icons, or by clicking the blue plus sign near the top left.

_images/REPL.png

Fig. 2.8 “Normal” interactive Notebook#

_images/CLI.png

Fig. 2.9 Unix/Linux Command Line#