Remote computing at the Unity cluster.¶

Use the documentation to supplement this guide

https://docs.unity.uri.edu/

Objectives¶

Learn basic unix shell navigation and commands.
Set up computing environment for Python on the Unity cluster.
Use file transfer to upload/download files from the remote environment, including Python (.py) scripts and image files (.jpeg, .png, .tif).
Practice basic tasks at the Unix command line.
Set up the environment for a parallel computing exercise

+We will compute the NDVI Normalized Distributed Vegetation Index using LANDSAT satellite imagery from the bands. Check out the details at the above link.

Step 1: Log in to Unity.¶

Your user account and password was created when you requested an account at unity.uri.edu.

These should be something like:

Name: first last name
Login: firstinitiallastname_uri
Password: whatever_you_specified

To access the HPC, we will use the Unity On Demand service via web browser.

In your browser type in:
```
$ https://ood.unity.rc.umass.edu/
```
Choose your institution and follow the steps to log in.

If this connection is successful, then you can proceed to setting up your Python environment in the next step.

NOTE: The Oscar Unity environment uses PyModules to let the user customize their work environment and even install software via conda or pip. Type ]$ module available to see what's available.

Step 2: Set up the working environment.¶

[Do this only once] Once you have loaded the Anaconda module, you need to install the Python libraries that we will use. This command creates an Anaconda programming environment called your_env_here. You will need to activate this environment every time you log in.
```
$ module load conda/latest
$ conda create -n [your_env_here] python=3.9 dask mpi4py scikit-image matlplotlib
$ conda activate [your_env_here] # Activate your environment before installing landsatexplore using pip.
$ pip install landsatxplore
```

[Do this every time you log in] Unity uses modules; modules can be loaded with the load command. You will load them every time you log in.
```
$ module load conda/latest
$ source activate [your_env_here]
```
At the end of the session when you log out, you can enter source deactivate to end the Anaconda environment.

RECAP:

The next time you log in, you won't have to carry out the installation steps. you will just execute the following command sequence.

$ module load conda/latest
$ source activate [your_env_here]

Step 3: Follow this Unix tutorial on navigating around the Oscar directories, creating files, etc.¶

Also check out Unix shell commands to better understand how to use the command line tools. https://maker.pro/linux/tutorial/basic-linux-commands-for-beginners. You will likely need to use the Unix commands sudo, cd, rm, mkdir, ssh, sftp, and nano.

Step 4: Transfer your .py files by OODs web-based file manager.¶

Choose Files-->Home Directory from the top menu in Unity Open On Demand.

If you haven't already, make a directory in your Unity home folder entitled ocg404. This directory should be inside your /home/your_username.

You can do this at the Shell:

]$ mkdir ocg404

Or you can create the folder from within the OOD file manager that you opened above with Files-->Home Directory.

Now transfer your files to the ocg404 directory on Unity OOD.

Upload landsatexplore.py

Step 5: Use a text editor to modify landsatexplore.py and upload via the Unity OOD file manager.¶

landsat_exploreRI.py uses the landsatxplore API to find and download Landsat images, based on the time and geographic criteria you give it.
To use the API, you will need to create a USGS EarthExplorer account.
Fllow the prompts in landsat_exploreRI.py and add the missing code to landsat_exploreRI.py.
You will download and graph at least one landsat image.
Note, that because we don't have a graphics connection to Unity, you will need to save the file as a .png or .jpeg for downloading by ftp.
Run landsat_exploreRI.py on the Unity HPC at the command line using the command below:

]$ python landsat_exploreRI.py

What to turn in (in-class assignment part 2):¶

Download the image you created and submit that as this week's in_class assignment, along with your modified version of landsat_exploreRI.py.

In [ ]: