Python SciServer

Download a folder with 2 Python notebooks

Direct link to github repository containing DEMO_Getdata_SciServer.ipynb and DEMO_Getcutout_SciServer.ipynb files, as well as other codes being run on the JHTDB backend.

Overview: The DEMO_Getdata_SciServer Python code illustrates sampling velocities on arrays of points (a 2D plane, a 3D volume, random sets of points, and a time-history at a single point). The DEMO_Getcutout_SciServer samples raw data on the grid points and timesteps (no interpolation or differentiation). The sample array is a 3D rectangular prism. Users can execute the Python code on SciServer and it will query the database during execution.

How to use SciServer

SciServer is a cloud-based data-driven cluster, of The Institute for Data Intensive Engineering and Science (IDIES) at Johns Hopkins University. Users get the advantages of more reliable and faster data access since the SciServer is directly connected to JHTDB through a 10 Gigabit ethernet connection. SciServer provides containers with the "giverny" library pre-installed.

To use giverny through SciServer:

1. Login to SciServer (may need to create a new account first).

2. Click on *Compute* and then *Create container* (You could also run jobs in batch mode, by selecting Compute Jobs).

3. Type in *Container name*, select *SciServer Essentials 4.0* in*Compute Image*, mark *Turbulence (ceph)* and *Turbulence Windfarm (ceph)* in *Data volumes*, and then click on *Create*.

4. Click on the container you just created, then you could start using giverny with IPython Notebook.