A few instructions to start working with impresso data exports in the context of the HUM-475 class.
Noto is the EPFL’s JupyterLab centralized platform for education which allows teachers and students to use Jupyter notebooks online without having to install anything on their computer.
- clone this repository
git clone https://github.com/impresso/epfl-shs-hum-475.git - go to https://noto.epfl.ch/ and login with your gaspar credentials.
- upload the example notebook and your data.
- Download Anaconda in order to get the Conda environement manager.
- Familiarize yourself with Conda
- Open a terminal, go to your working repository and create an environement:
conda create -n NAME python=3.6where NAME is the name you want to give to the environement (e.g. digital-history) - Activate it:
source activate NAME - install dependencies with
pip install -r requirements.txt
Useful commands (and more info here):
conda info --envs => list your environments
source deactivate => deactivate an env
conda remove --name NAME --all => remove environment 'NAME'
What it is: see this tutorial
Conda already installs by default Jupyter when you create an environment.
To launch a notebook, just execute this in your activated env:
jupyter notebook
We've put a jupyter notebook in this repo (explore_export.ipynb) where you can get an idea where to start.
If you want to use Iramuteq, you will have to isolate the textual parts and print them as specified here.
Contribution from previous years' groups who developed a script to process the data before importing into Iramuteq.
- Projet sur l'OPEP: scripts, rapport.
- Projet sur le secret bancaire: scripts, rapport.
- Projet sur la relation americano-soviétique: scripts, rapport.
Ouvrir R et installer manuellement les librairies suivantes:
install.packages('ca')
install.packages('textometry')
install.packages('irlba')
install.packages('wordcloud')
install.packages('proxy')
install.packages('igraph')
install.packages('ape')
install.packages('gee')
install.packages('rgl')