YODA principles

docs/basics/101-122-intro.rst

@@ -0,0 +1,23 @@
+.. _intromidterm:
+
+A Data Analysis Project with DataLad
+------------------------------------
+
+
+Time flies and the semester rapidly approaches the midterms.
+In DataLad-101, students are not given an exam -- instead, they are
+asked to complete and submit a data analysis project with DataLad.
+
+The lecturer hands out the requirements: The projects needs to
+
+- be prepared in the form of a DataLad dataset
+- needs to contain a data analysis performed with Python tools
+- should incorporate DataLad whenever possible (data retrieval, publication,
+  script execution, general version control) and
+- needs to comply to the YODA principles
+
+Luckily, the midterms are only in a couple of weeks, and a lot of the
+requirements of the project will be taught in the upcoming sessions.
+Therefore, there's little you can do to prepare for the midterm
+than to be extra attentive on the next lectures on the YODA
+principles and DataLads Python API.

docs/basics/101-124-summary.rst

@@ -0,0 +1,53 @@
+.. _summary_yoda:
+
+Summary: YODA principles
+------------------------
+
+The YODA principles are a small set of guidelines that can make a huge
+difference towards reproducibility, comprehensibility, and transparency
+in a data analysis project.
+
+These standards are not complex -- quite the opposite, they are very
+intuitive. They structure essential components of a data analysis project --
+data, code, computational environments, and lastly also the results --
+in a modular and practical way, and use basic principles and commands
+of DataLad you are already familiar with.
+
+There are many advantages to this organization of contents.
+
+- Having input data as independent dataset(s) that are not influenced (only
+  consumed) by an analysis allows for a modular reuse of pure data datasets,
+  and does not conflate the data of an analysis with the results or the code.
+
+- Keeping code within an independent, version-controlled directory, but as a part
+  of the analysis dataset, makes sharing code easy and transparent, and helps
+  to keep directories neat and organized. Moreover,
+  with the data as subdatasets, data and code can be automatically shared together.
+
+- Including the computational environment into an analysis dataset encapsulates
+  software and software versions, and thus prevents re-computation failures
+  (or sudden differences in the results) once
+  software is updated, and software conflicts arising on different machines
+  than the one the analysis was originally conducted on.
+
+- Having all of these components as part of a DataLad dataset allows version
+  controlling all pieces within the analysis regardless of their size, and
+  generates provenance for everything, especially if you make use of the tools
+  that DataLad provides.
+
+- The yoda procedure is a good starting point to build your next data analysis
+  project up on.
+
+Now what can I do with it?
+^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+Using tools that DataLad provides you are able to make the most out of
+your data analysis project. The YODA principles are a guide to accompany
+you on your path to reproducibility.
+
+What should have become clear in this section is that you are already
+equipped with enough DataLad tools and knowledge that complying to these
+standards will feel completely natural and effortless in your next analysis
+project.
+The next section will add to your existing skills by demonstrating how to
+use DataLad also within Python scripts.

docs/contents.rst.inc

@@ -93,6 +93,18 @@ Help yourself
 
    basics/101-135-help
 
+#########################
+Data analyses in datasets
+#########################
+
+.. toctree::
+   :maxdepth: 1
+   :caption: Organizational principles and best practices for data analyses
+
+   basics/101-122-intro
+   basics/101-123-yoda
+   basics/101-124-summary
+
 #########
 Use Cases
 #########