Laptop configuration for the training

The training will be conducted using X2Go-servers available in the TCE. Instructions on how to install X2Go on your laptop can be found in DM Tool document EUM/RSP/VWG/18/1018393.

  1. You should have X2Go already installed on your laptop, so go ahead and start it up using your TCE credentials
  2. Open a new console from the Start menu -> System Tools -> Konsole (font size can be made bigger with Ctrl +)

Installation and use of conda

Installation of miniconda3

Type the following commands to the terminal.

  1. wget -O ~/
  2. bash ~/ -b -p $HOME/miniconda3
  3. echo 'export PATH="$PATH:$HOME/miniconda3/bin"' >> ~/.profile.local
  4. source ~/.profile.local

Create a new conda environment for the training

The name of the environment can be chosen freely, for the sake of simplicity we will use name troll for this training.

  1. conda create -n troll python=3.6

Activate the troll environment with a command

  1. source activate troll

You should have an indication in your terminal prompt that the conda environment is active, e.g.: (troll) [email protected]:~ $. When the conda environment is active, all the installations after this point will go only to this specific environment. The environment can be deactivated with source deactivate-command.

Installation of required libraries and modules

Install the required libraries and modules with this monster one-liner:

  1. conda install -c conda-forge gdal satpy jupyter ffmpeg "imageio<2.5" imageio-ffmpeg shapely cartopy scipy rasterio mysql-connector-python matplotlib cartopy scipy pycoast pyorbital

Apply updates to packages

PyTroll packages are constantly being developed. In order to carry out some the exercises, we need to get an updated version of one of the readers.

  1. pip install git+
  2. cp /tcenas/scratch/pytroll/ ~/miniconda3/envs/troll/lib/python3.6/site-packages/satpy/readers/

Using jupyter-notebook

Copy the exercises

  1. mkdir ~/pytroll_training
  2. cp /tcenas/scratch/pytroll/exercises/* ~/pytroll_training
  3. cd ~/pytroll_training

Starting headless jupyter notebook server

In order to get the best performance for the exercises, we'll start jupyter notebook with a headless server, i.e., we run the notebook on the X2Go-server, but connect to it from your own laptops.

Find out the IP-address of your server, and start jupyter notebook with --ip and --no-browser options

  1. /sbin/ifconfig
  2. jupyter notebook --ip=<ip_address> --no-browser

Copy and paste the given URL to your laptops Google Chrome browser

In [ ]:
import satpy
print (satpy.__file__)