Page 293 - Contributed Paper Session (CPS) - Volume 4

CPS2230 Braden P. et al.
there is any correlation among the selected statistical variables and whether
or not the data are also spatially correlated, as reflected in the filled-in
shapefiles found within the spatial panel.
In Section 2 of this article, we describe the Mapshaper and ‘rmapshaper’
software that form the basis for creating meaningful shapefiles for further use
in LMplots. The motivation why such modified shapefiles are needed is given
in Section 3. The methods used to modify shapefiles from within R are
summarized in Section 4. An example for Canada follows in Section 5. We
finish with a discussion and conclusion in Sections 6 and 7, respectively.

2. Methodology
Mapshaper, and its web browser version mapshaper.org [2], is a software
tool introduced to the public by Matthew Bloch and Mark Harrower in 2006
as an open-source tool to modify and simplify shapefiles. Before the
introduction of Mapshaper, there were a few software programs that allowed
users to modify shapefiles as needed. However, the real strength of
Mapshaper is that it was the first, free software program that provided a
WYSIWYG (what you see is what you get) approach for users that are not
trained or prepared in modifying shapefiles.
While most of the functionality of the Mapshaper software is still done
manually through a command line, the syntax of the commands is easy to
understand even for beginners. Further, with the shapefile being updated in
close to real-time, users are able to see exactly how each modification would
appear in the final product.
Andy Teucher, in 2016, brought even more accessibility to the modifying
and simplifying of shapefiles when he released an R package titled
‘rmapshaper’ [7]. ‘rmapshaper’, at its core, provides an R wrapper to the
functionality provided within mapshaper.org. As ‘rmapshaper’ is still fairly new,
not every command in Mapshaper has been directly translated into a
standalone R function. While ‘rmapshaper’ is still being updated to include
more commands as R functions, Andy Teucher has provided access to the
entirety of Mapshaper commands through the inclusion of the command
apply_ mapshaper_commmands(). While not necessarily ideal, this still does
allow us to use all of the commands to tailor our shapefiles to be exactly what
we need.

3. Motivation
While shape files for various countries and other regions of interest are
readily available through the internet, e.g., via the Database of Global
Administrative Areas (GADM) [1], plotting shapefiles in their “raw” format,
especially coastal regions, is a time consuming and computationally expensive
process, especially when plotting several repeated plots. Further, many of the

282 | I S I W S C 2 0 1 9