docs: add ref to public training set data

docs/index.rst

@@ -13,6 +13,7 @@ Documentation
 
    sec_getting_started
    sec_cli
+   sec_rating_workflow
    sec_examples
    sec_code_reference
 

docs/nanite.bib

@@ -1,3 +1,12 @@
 % Encoding: UTF-8
 
+@Article{zef18,
+  author  = {Paul Müller and Stephanie Möllmert and Jochen Guck},
+  title   = {{Atomic force microscopy indentation data of zebrafish spinal cord sections}},
+  journal = {Figshare},
+  year    = {2018},
+  month   = {11},
+  doi     = {10.6084/m9.figshare.7297202.v1},
+}
+
 @Comment{jabref-meta: databaseType:bibtex;}

docs/sec_rating_workflow.rst

@@ -0,0 +1,42 @@
+.. _sec_rating:
+
+===============
+Rating workflow
+===============
+One of the main aims of nanite is to simplify data analysis by sorting out
+bad curves automatically based on a user defined rating scheme.
+Nanite allows to automate the rating process using machine learning,
+based on `scikit-learn <http://scikit-learn.org/>`_.
+In short, an estimator is trained with a sample dataset that was manually
+rated by a user. This estimator is then applied to new data and, in an
+optimal scenario, reproduces the rating scheme that the user intended
+when he rated the training dataset.
+
+Nanite already comes with a default training set that is based on AFM
+data recorded for zebrafish spinal cord sections, called `zef18`.
+The original dataset is available on figshare :cite:`zef18`.
+Download links:
+(SHA256 sum: 63d89a8aa911a255fb4597b2c1801e30ea14810feef1bb42c11ef10f02a1d055).
+
+- https://ndownloader.figshare.com/files/13481393
+
+With nanite, you can also create your own training set. The required steps
+to do so are described in the following.
+
+
+Rating experimental data manually
+=================================
+- ref to available models 
+- ref to fitting guide
+- nanite-setup-profile
+- nanite-rate
+
+
+Generating the training set
+===========================
+- nanite-generate-trainining-set
+
+
+Applying the training set
+=========================
+- set file system location of training set in rate_quality