inseefrlab / lockdown-maps-r Goto Github PK

For now, there are two means to (re)produce the dataviz: using the targets package or the FranceMNO.R script. I suspect that can be confusing for an external reader. Moreover, the dataviz generation process is not (yet) documented.

Do you want to keep both methods or just one (which one in that case)?

This link to the data works: https://github.com/InseeFrLab/lockdown-maps-R/blob/master/data/MouvementsPopulation_confinement2020.csv. However the link in the readme (Dataviz Data, as published in April 2021: allows re-use by disseminating flows, is broken.

Hi @milena-git and @mtennekes,

@MelineeTS and me are preparing the open sourcing of this project on behalf of the Insee.

I open this issue to discuss about the FOSS license.

@mtennekes: in 🇫🇷 , the FOSS licenses that we can use are limited by the law.
We can use one of the following licenses.

For the code, the permissive ones are:

• MIT
• Apache-2.0
• BSD-2-Clause
• BSD-3-Clause
• CECILL-B

For the code, the ones with a copyleft are:

• CECILL-2.1
• CECILL-C
• GPL-3.0-or-later
• LGPL-3.0-or-later
• AGPL-3.0-or-later
• MPL-2.0

For the datasets and the publications, we can use:

• ODbL-1.0
• etalab-2.0 (this is a 🇫🇷 governmental permissive license)

Even if the EUPL-1.2 is not listed here, we can consider it for the code.

The Insee encourages the use of the MIT licence but any of the above mentioned licenses would be correct.

@mtennekes , would you agree to release this conjoint code under the MIT license?

This link to TCRD_021.xls file seems to be broken

Got the following error on Windows:

Error in if (any(abs(object@deg) > 360)) return("abs(degree) > 360") else if (any(object@WS &  : 
  valeur manquante là où TRUE / FALSE est requis

How to reproduce?

library(tidyverse)
source("R/maps_utils.R")
departement_centroids()

The origin of this bug comes from the use of source() and these lines:

The problem with source() is that the default value for the encoding parameter is getOption("encoding") which is "native.enc" on Windows.
Here, the files are encoded in UTF-8 and the ° character is used. When parsed as CP-1252, this character is wrong.

There are two options to solve this bug:

• always use the encoding value for source() calls: source(..., encoding = "UTF-8")
• replace "°" with "\ub0"

@milena-git, I can do the PR if you want.

Finding how to cite these dataviz is not yet straightforward.
We could add an entry in the README file ("Cite this work") and/or a CITATION file.

A README file is needed in the project

For now, the HTML files are commited in the html directory. These files are huge.
Moreover, commiting artifacts is not a best practice.
I suggest to produce these files with GHA and make them available as artifacts.

The html outputs of the targets (locally and build in github actions artefacts) are not synchronized anymore. Seems also the case for the last version of FranceMNO.R
@mtennekes any idea?

I created a new and public R package: https://github.com/mtennekes/donutmaps

• It (only) contains the function back_donuts.
• One change w.r.t. the old function (that is also included in this repo): I've added the parameter var, which contains the variable name of flow values. In the old case (and in your case), it should be "flow"

@milena-git The old private repo donut_maps still exists, but I'll have to remove bake_donuts there as well, and probably rename this repo.