Merge branch 'feature/cloudcom2016' of github.com:Falkor/RR-tutorials…

… into feature/cloudcom2016

slides/2016/cloudcom2016/src/_challenges.md

@@ -0,0 +1,55 @@
+### Reproducibility axes
+
+\begin{textblock}{0.4}(0.6,0.22)
+  \imgw{}{RR-3-axis.png}
+\end{textblock}
+
+* Always keep track of:
+    - your \emph{methodology}
+    - your \emph{code}
+    - your (input) \emph{data}
+
+* Can you later come back and:
+    - reproduce you experiment
+    - including its environment
+    - ... and obtain the same results?
+
+\pause
+\vspace{1.8em}
+
+* If \emph{not}, then \emph{now} is the best time to start
+    - documenting your processes
+    - describing your environment (software \emph{and hardware}!)
+    - versioning and tagging your code and data
+    - (... and keep backups of it all)
+
+### Reproducibility levels
+
+\begin{textblock}{0.4}(0.65,0.11)
+  \smartdiagramset{
+   planet size=2.5cm,
+   satellite size=2.3cm,
+   distance planet-text=0,
+   distance planet-satellite=3cm, 
+   /tikz/connection planet satellite/.append style={<-}
+  }   
+  \centering\scalebox{0.5}{\smartdiagram[constellation diagram]{\emph{Research},
+  \hspace*{-0.3em}\mbox{Reviewable}, \hspace*{-0.3em}\mbox{Replicable}, 
+  \hspace*{-0.5em}\mbox{Confirmable}, \hspace*{-0.2em}\mbox{Auditable}, \hspace*{-0.7em}\mbox{Reproducible}}}
+\end{textblock}
+
+\emph{Is your research}\footnote{ICERM Report 2013: "Reproducibility in Computational and Experimental Mathematics"}:
+
+* reviewable
+    - desc. of your methods can be independently assessed?
+* replicable
+    - are the tools available to duplicate the results?
+* confirmable
+    - can the main conclusions be attained independently of your tools?
+* auditable
+    - do you have records such that your research can be later defended?
+    - ... or differences between independent confirmations resolved?
+* \emph{open or reproducible}, such that
+    - the procedures can be fully audited \emph{and}
+    - the results can be replicated or independently reproduced \emph{and}
+    - the results can be extended or the method applied to new problems

slides/2016/cloudcom2016/src/_me.md

@@ -10,7 +10,7 @@
 * _2003 -- 2007_: PhD between INP Grenoble \& UL
      - \textit{Security in Large Scale Distributed Systems: \newline Authentication and Result Checking}
 * _2007 -- now_: \textit{Research Associate} at [UL](http://www.uni.lu)
-    - Part of the [PCOG Team](http://pcog.uni.lu) lead by Prof. P. Bouvry
+    - Part of the [PCOG Team](http://pcog.uni.lu) led by Prof. P. Bouvry
     - Manager of the [UL High Performance Computing Facility](http://hpc.uni.lu)
          * $\simeq$ 197 **T**Flops (2017), 5.844 **P**B, 4 sysadmins
 

slides/2016/cloudcom2016/src/_vagrant.md

@@ -54,7 +54,7 @@ $> brew cask install vagrant-manager # cf http://vagrantmanager.com/
      - `vagrant up`
 * Keep the number of VMs under control
      - All configuration in `VagrantFile`
-* _Reproducability_
+* _Reproducibility_
      - Identical environment in development and production
 * _Portability_
      - _avoid_ sharing 4 GB VM disks images

slides/2016/cloudcom2016/src/tutorial-RR.tex

@@ -157,8 +157,8 @@ \subsection{Mastering your [reproducible] environment}
 % % Automatic testing
 % % - Bats, Travis, gitlab-CI, codeclimate....
 
-% % Open Challenges in Numerical Repoducibility
-
+% % Open Challenges
+\input{_challenges.md}
 
 
 % % ======================== END =========================