Open-data is often not in an appropriate format. When it is the case, hackers like ourselves need some efforts to clean it. Each folder contains:
- a README explaining waht dataset is being cleaned, what are the outputs etc
 - a makefile to download the original data
 - scripts to run the moulinette
 
Don't hesitate to contribute.
- 
insee-iris converts all the Insee IRIS shapefiles encoded in Lambert 93 into a single geojson file encoded with longitude/latitude.
 - 
insee will download all the datasets of the INSEE at the iris level and cleanly aggregate them in on single csv file.
 - 
weather_forecast to have four days of weather prediction (only for Switzerland, France and Belgium) asked for a city.
 - 
caf will download all the datasets from caf open-data at the city level and cleanly aggregate them in on single csv file.
 - 
impot_locaux will download all files from elements of local taxation at city level.
 
We made a workshop to show how we can visualize and query these datasets with elastic search and kibana.