All posts by Ewen Gallic


Morning talks – COVEA


I am going to Paris today to attend a meeting this morning with the people from the Actinfo Chair that I am now part of for the duration of my post-doc with Arthur Charpentier.

I will present the research on genealogy using collaborative data we have been working on this summer. This will be the occation to speak about what we plan to do with Arthur regarding those data in the near future.

Olivier Wintenberger will also share with us his recent research.


GEOMED 2017 conference in Porto (Portugal)

Porto 2017

This week I will attend the GEOMED2017 conference in Porto, in Portugal. Researchers from different horizons will be gathering here to attend to some talks about spatial statistics, spatial epidemiology and public health. This will be the perfect occasion for me to learn more on those subjects during the three days of the conferences. I will also attend a workshop given by Lee Duncan from Glasgow University on modelling spatial data in R with the package CARBayes).


Where’s Waldo? Here he is!

Yesterday, I came across a nice article untitled “Here’s Waldo: Computing the optimal search strategy for finding Waldo“, written by Randal S. Olson. I used the data he shared to apply a correction to the kernel density estimation of Waldo’s location.

In this article, Randal explains that he has devoted some time to try to compute the optimal search strategy for finding Waldo. To that end, he has used some machine learning techniques.
From an image provided by Slate (Here’s Waldo, 2013, by Ben Blatt), Randal S. Olson retrieved the coordinates of 68 different locations of Waldo, and kindly shared the data afterwards.



Qui va gagner la Nouvelle Star ? À vos votes Twitter !

Cumul des votes par candidat au cours de la soirée

La chaîne de télévision D8 diffuse pour la troisième année consécutive l’émission Nouvelle Star. Si ma mémoire est bonne, je n’avais pas suivi ce programme depuis la saison 2006-2007, mais j’avoue m’être pris au jeu cette année, d’autant plus que la production propose une nouveauté par rapport aux saisons précédentes : les votes par Twitter !

Jeudi dernier, le 12/02, lors du quatrième prime, je me suis diverti avec l’API de Twitter pour récupérer les tweets postés pendant l’émission.

Les résultats de ma petite analyse laissent penser que les votes par SMS et par téléphone pratiqués durant l’émission ne sont pas négligeable, et ne reflètent pas exactement ce qui se dit sur le réseau social Twitter.

Aussi, après un bref rappel du déroulement des primes, ce billet présentera quelques observations réalisées autour des tweets extraits pendant l’émission de jeudi. Une troisième partie s’adressant à ceux désirant récupérer eux-même les tweets et effectuer leur propre analyse, fournira quelques codes R.


French dates in R – From Mac OS to Windows

Station Clock - By derivative work: FischX (talk) Station_Clock.jpg: User:JuergenG, User:AlMare (Station_Clock.jpg) [CC-BY-SA-2.5-2.0-1.0 ( or CC-BY-SA-2.5-2.0-1.0 (], via Wikimedia Commons
The other day during my R lecture, something I did not expect happen… I should have known, or at least guess that it would not work… But I wanted to believe it would. When I gave examples on how to read both English and French formatted dates in R, what was working perfectly on Mac OS X did not on Windows Seven. The example was the following:

d_char_fr <- "Mer 04 Fév 2015"
d_char <- "Wed 04 Feb 2015"
as.Date(d_char_fr, format = c("%a %d %b %Y"))
as.Date(d_char, format = c("%a %d %b %Y"))

Coordinates of “hot-spots”

Last week I uploaded a note about the paper @freakonometrics and I have submitted again after we added quite a few revisions. We have already made available most of the R code to reproduce our applications (on this GitHub repository). However, we did not share our method to obtain the coordinates of the “hot-spots” areas