EXERCISE for Remote Sensing Course

This repository contains the exercices for the EPFL Remote Sensing Course. There are three options for executing the code:

1. On EPFL's virtual machines, available in the exercise classroom;
2. On your personal laptops;
3. Via a Noto link.

We recommend using options 1 or 2 for pedagogic purposes (see section 3 for more details on why).

1. Instructions for using EPFL's VM

Please select the ENAC-SSIE-Ubuntu-20-04 Virtual Desktop Infrastructure (VDI) and then follow these steps:

1. Download the RS2026 GitHub repository

2. Unzip the RS2026-main.zip file and move the RS2026 directory within the /home/<your_username>/Desktop/myfiles/ directory. If your data are saved within the /myfiles directory, they will be available the next time you reconnect to the VDI.

3. Open a terminal and activate the lte environment with:

micromamba activate lte

4. Then create the lte ipykernel for Jupyter Notebook with:

python -m ipykernel install --user --name=lte

5. Launch the Jupyter Notebook interface with jupyter notebook and open the Exercise_6.ipynb or Exercise_7.ipynb file within the RS2026 directory.

6. To execute correctly the Jupyter Notebook, in the top menu bar select Kernel > Change Kernel... and switch the kernel from Python 3 (ipykernel) to lte.

2. Instructions for using your own computer

Alternatively, you can clone the RS2026 repository on your laptop and install the required environment using conda/mamba or micromamba:

1. Go to the directory where you want to clone the repository. As an example:

cd /home/ghiggi/courses

2. Clone this repository:

git clone git@github.com:ltelab/RS2026.git
cd RS2026

3. Install the dependencies using conda:

micromamba env create -f environment.yml

4. Activate the lte conda environment:

micromamba activate lte

4. Create the lte Jupyter Notebook environment with:

python -m ipykernel install --user --name=lte

5. Launch the Jupyter Notebook interface with jupyter notebook and open the Exercise_6.ipynb or Exercise_7.ipynb file within the RS2026 directory.

6. To execute correctly the Jupyter Notebook, in the top menu bar select Kernel > Change Kernel... and switch the kernel from Python 3 (ipykernel) to lte.

If you want to use the VScode interface for executing the Jupyter notebook, execute steps 1 to 5, then install Python and Jupyter extensions in VScode. Navigate to the repository using VScode's file selector, then open the .ipynb files. Click the kernel selector on the top-right corner of the notebook and select the lte environment.

Note that the installation of the dependencies on your laptop might cause conflicts; the latest version of the required packages can be installed using the following command:

conda install numpy pandas xarray dask rasterio rioxarray scikit-learn matplotlib-base seaborn colorcet pywavelets pillow jupyter

In case you encounter such issues and cannot fix them, please contact the TA team.

3. Instructions for using Noto

To run the exercise, you can also use noto.epfl.ch JuypterLab service (more details here). This allows you to run the code directly without having to do any environment setup. We do not recommend this option, as we think that setting up an environment is a good thing to learn; but we provide this option as a backup if you encounter issues with option 1 and 2.

1. Click here to access the Noto environements

2. Once you have access to the EPFL Noto platform, navigate in the filesystem (left panel) to the right exercise / notebook and click on the ipynb file to start the notebook.

---

And now ... happy coding :-)