I've been searching for a way to automatically calculate breaks to reclassify a raster using QGIS or another open source software, in a similar way it can be done in ArcGIS (Slice tool). I've searched and I found some related questions like QGIS Natural Breaks with raster file, Apply “natural breaks” classification to a raster symbology in QGIS? and calculation of breaks for a quantile classification in GRASS (Raster style based on percentile QGIS). The questions related to natural breaks didn´t have an answer. I've also checked some R packages like classInt, but it seems there are problems for large rasters, as it's shown in this post in the r-sig-geo forum (Jenks classification for raster representation).
I've recently been reading the book QGIS 2 Cookbook (a good book by the way) and they say that the SAGA Cluster analysis for grids algorithm could be used to reclassify a raster. I haven´t found many examples of using this algorithm for raster reclassification, apart from the one in the book Geomorphometry, in the chapter related to Geomorphometry with SAGA and in this presentation (Digital Soil Mapping in SAGA GIS).
What I am trying to do is to convert a quantitative map in a qualitative map, let's say I have a fire risk raster and I want to reclassify it to low, medium, high and very high classes.
Is this algorithm suitable for this kind of raster reclassification?
Are there other ways of automatically calculate the classes breaks using QGIS or another open source software with the Jenks Natural Breaks algorithm or other algorithms different from Equal Interval or Continuous?
Cluster analysis for grids(now calledK-Means Clustering for Grids) tool, it may not be the tool you are looking for because it takes into account neighboring grid cells. – Kazuhito Sep 08 '17 at 19:45