aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
authorMohammad Akhlaghi <mohammad@akhlaghi.org>2019-07-08 15:58:19 +0100
committerMohammad Akhlaghi <mohammad@akhlaghi.org>2019-07-08 15:58:19 +0100
commit9f8cc748c174c11933663383169e6c9eb3b10c57 (patch)
tree93a6507930c3936ecd426a088c505e1882f4fc62
parent8ebb7848bf95d07e7162f46387ff4702f685da93 (diff)
Minor corrections before presentation
Before the presentation, some minor corrections were made to make the points more clear.
-rw-r--r--reproducible-paper.tex76
1 files changed, 63 insertions, 13 deletions
diff --git a/reproducible-paper.tex b/reproducible-paper.tex
index 475b60d..fb9f732 100644
--- a/reproducible-paper.tex
+++ b/reproducible-paper.tex
@@ -12,8 +12,9 @@
(\textcolor{blue}{\href{https://arxiv.org/abs/#1}{arXiv:#1}})}}
%% Set the title
-\title{\LARGE Digging out the low-surface-brightness universe\\
- \small understanding and constraining the limits of methods used}
+\title{\LARGE Digging out the low-surface-brightness
+ universe,\\ \textbf{BIG} Data, \textbf{BIG} responsibility
+ \\ {\normalsize (a practical solution for reproducibile science)}}
%% Set the author
\author{Mohammad Akhlaghi\\\vspace{2mm}\footnotesize Instituto de
@@ -91,20 +92,20 @@
+
+
\newcommand{\imgtdir}{/home/mohammad/documents/personal/professional/astronomy/talks/in-prep/images}
\begin{frame}{Reproducibility is critically important for low surface brightness astronomy}
\small
\begin{columns}
- \column{6cm} Example: Outer surface brighntess of M51 in a
+ \column{6cm} Example: Outer surface brightness of M51 in a
\alert{single-exposure} SDSS image, using NoiseChisel.
\vspace{2mm}
\begin{itemize}
\setlength\itemsep{3mm}
\item Outer wing detected to \alert{$\rm{S/N}=1/20$} or
- \alert{$30.06$ mag/arcsec$^2$}.
- \item Deep/orange image from Watkins+2015
- (\textcolor{blue}{\href{https://arxiv.org/abs/1501.04599}{arXiv:1501.04599}}).
+ \alert{$\sim30$ mag/arcsec$^2$}.
\item
\textcolor{blue}{\href{https://www.gnu.org/software/gnuastro/manual/html_node/Detecting-large-extended-targets.html}{Complete
tutorial}} in manual fully describes how to derive/reproduce
@@ -113,10 +114,16 @@
\item \alert{Run-time} options/configuration.
\item Steps \alert{before/after} NoiseChisel.
\end{itemize}
- \item Default settings will \alert{not} give this result.
+ \item Deep/orange image from Watkins+2015
+ (\textcolor{blue}{\href{https://arxiv.org/abs/1501.04599}{arXiv:1501.04599}}).
+ \item Therefore:
+ \begin{itemize}
+ \item Default settings \alert{not enough}.
+ \item Result not \alert{just from NoiseChisel}.
+ \end{itemize}
\end{itemize}
- \vspace{2mm}
+ \vspace{1mm}
\begin{tcolorbox}[boxsep=0pt,left=1mm,right=1mm,top=1mm,bottom=1mm]
Simply reporting in your paper that ``\emph{we used
NoiseChisel}'' is \alert{not enough} to reproduce, understand,
@@ -135,6 +142,7 @@
+
\begin{frame}{Reproducibility crisis in the sciences/astronomy}
\begin{tcolorbox}[title=Snakes on a Spaceship -- An Overview of Python in Heliophysics]
\small ``...\alert{inadequate analysis descriptions} and loss of
@@ -181,11 +189,12 @@
\rule{0.5\linewidth}{1pt}
\end{center}
\begin{itemize}
- \setlength\itemsep{1.5em}
+ \setlength\itemsep{0.5em}
\item Involves data \alert{collection}.
\item Inherently includes \alert{measurements errors}\\(can
never be exactly reproduced).
\item Example: Raw telescope image/spectra.
+ \item \alert{\textbf{NOT DISCUSSED HERE (Nacho's talk).}}
\end{itemize}
\vspace{3.5mm}
@@ -202,15 +211,15 @@
\rule{0.5\linewidth}{1pt}
\end{center}
\begin{itemize}
- \setlength\itemsep{1.5em}
+ \setlength\itemsep{1em}
\item Involves data \alert{analysis}, or simulations.
\item Starts \alert{after} data is collected/digitized.
\item Example: $2+2=4$ (i.e., sum of datasets).
- \item \textbf{\textcolor{green!50!black}{The type discussed here.}}
+ \item \textbf{\textcolor{green!50!black}{DISCUSSED HERE.}}
\end{itemize}
\centering
- \vspace{1mm}
+ \vspace{3mm}
\includegraphics[width=0.8\linewidth]{img/digital-tunnel.jpg}\\
\vspace{-0.6mm}
\tiny \href{https://tsongas.com/newsletter_articles/the-new-electronic-version-of-the-advantage/digital-tunnel-wallpaper/}{https://tsongas.com}
@@ -304,7 +313,7 @@
\vspace{-1mm}
\begin{tcolorbox}[boxsep=0pt,left=1mm,right=1mm,top=1mm,bottom=1mm]
\small Data analysis [...] is a \alert{human
- behaviour}. Researchers who hunt hard enough will turn up a
+ behavior}. Researchers who hunt hard enough will turn up a
result that fits statistical criteria, but their
\alert{discovery} will probably be a \alert{false positive}.
@@ -855,6 +864,47 @@
+ \begin{frame}{Future prospects...}
+ \large Adoption of reproducibility by many researchers will enable
+ the following:
+
+ \vspace{1em}
+ \begin{itemize}
+ \setlength\itemsep{3mm}
+ \item A repository for education/training \textcolor{gray}{(PhD
+ students, or researchers in other fields)}.
+ \item Easy \alert{verification}/\alert{understanding} of other
+ research projects \textcolor{gray}{(when necessary)}.
+ \item Trivially \alert{test} different steps of others' work
+ \textcolor{gray}{(different configurations, software and etc)}.
+ \item Science can progress \alert{incrementally}
+ \textcolor{gray}{(shorter papers actually building on each
+ other!)}.
+ \item Applying \alert{machine learning} on reproducible research
+ projects will allow us to solve \alert{Johan's Big Data
+ Challenges} \textcolor{gray}{(from this morning)}:
+
+ \vspace{1em}
+ \begin{itemize}
+ \setlength\itemsep{2mm}
+ \item ``\emph{Extract the relevant parameters automatically}''.
+ \item ``\emph{Translate the science we will see this week to enormous
+ samples}''.
+ \item ``\emph{Believe the results when no one will have time to
+ reproduce}''.
+ \item ``\emph{Have confidence in results derived using machine
+ learning or AI}''.
+ \end{itemize}
+
+ \item \textcolor{green!70!black}{\textbf{Sixth paradigm of scientific
+ discovery?}}.
+ \end{itemize}
+ \end{frame}
+
+
+
+
+
\begin{frame}{GOOD NEWS: RDA adoption grant to IAC for this template}
\begin{center}
\includegraphics[width=3cm]{img/rda.png}\hspace{1cm}