March '23: The notebooks in this repo are no longer supported with the latest OceanSpy version. To see working, supported versions of these notebooks visit: https://github.com/hainegroup/Poseidon-share
Repository containing the tutorial notebook run during the Townhall Sesion
TH11 - Open Science Analysis of Petabyte Scale Ocean and Ocean-Atmosphere Models with Open Source Cloud Tools
How to:
You can download the contents of this repository into any machine with internet access and git
installed, by executing the following in the command line:
git clone https://github.com/hainegroup/OceanSciences2022.git
To reproduce the results of the tutorial notebooks, they need to be downloaded and executed within Sciserver to access the model data and correct (conda) environment. Furthermore, within Sciserver, the notebooks must run within a container created with the Oceanography 2.3
Compute image, along with access to either the Poseidon
(for llc4320) or Ocean Circulation
(for ECCO) data volumes.
The folder img/
contains static images used in the notebooks for model-data comparison. The folder needs to be added within the (local) directory as to where
the tutorial notebooks are stored, so that within the notebooks img/figure.png
sucessfully loads the static images.
To run the notebook :
- Make sure you create a container with
Oceanography 2.3
compute image along withOcean Circulation
data volume. You will be redirected into a jupyter lab environment. Navigate toTemporary/user/scratch
. This is where you will run the notebook. You can do this simply by clicking and opening these folders. If you're comfortable with working on a terminal , you will have to go to/scratch
using the following command
cd /home/idies/workspace/Temporary/your_username/scratch/
- You should have the notebook
LLC_90_Sciserver_tutorial.ipynb
along with the folders/img
and/test_data
to successfully reproduce the Ocean Sciences tutorial. If you are comfortable usinggit
you can clone this repo into your/scratch
folder. If you have no prior experience withgit
, simply download this repo and upload the folders into your jupyter lab environment on SciServer. The jupyter lab interface gives you the option to upload files and folders from your local system.
- Follow the same steps as for the LLC90 tutorial, except make sure that when you create your container, you additionally select the
Poseidon
data volume.