Improve User Documention

The following points could be illustrated in the example (/doc/source)

• How to use and create additional spectral indices
• How to use a few selected features, e.g., perform a classification with only the b2 and the NDVI of a Sentinel-2 image or time series
• How to write outputs
  • With/without the gapfilling
  • Select few features (e.g., R+IR)
  • how to re-use features resulting from "writeOutputs : True"
• How to use other images than Sentinel-2 (e.g., S1, Spot-6, Landsat-8, Pleiades)
• How to combine several images data sources (e.g., S2+S1, S2+VHR)
• How the sampling of the vector file is done and how it is possible to tune it: [Doing - @ArthurV ]
  • For the split between training/validation polygons [Doing - @ArthurV ]
  • For the selection of pixels insides the training polygons [Doing - @ArthurV ]
• how to check the integrity of the configuration file (e.g., comma separated list ...) and of the data (e.g., right projection, right folder ...)
• How to use IOTA² as a python library #121 (closed).
• Create a documentation table for ressources block name / steps name
• Documentation for dimensionnality reduction #162 (closed)
• Add information to use an interact with iota2 dev team #116 (closed)
• Add tutorial for each builder (classification / vectorization / features map)
• Add graphical pipeline #267