Source Code for the analysis published in: Rubio-Campillo, X., Ble, E., Pujol, À., Sala, R. and Tamba, R., 2022. A spatial connectivity approach to landscapes of conflict: Julius Caesar and the assault to Puig Ciutat (NE Iberian Peninsula). Journal of Archaeological Method and Theory, pp.1-31. The paper is fully available under an Open Access license from: https://doi.org/10.1007/s10816-022-09549-7
This analysis uses a diversity of open source packages including:
- QGIS 3.22 (cartography,sampling and spatial analysis)
- R and libraries ggplot2, raster, sf, gridExtra (statistics and data visualization)
- Circuistcape 5.6.0 via Julia (connectivity modelling)
- GDAL for command-based raster functionality
- Download the entire project dataset from https://doi.org/10.6084/m9.figshare.14995245.v3
- Clone or download this repository
- Extract the contents of the zip file within the root folder of the repository (it will create two new folders: 'data' and 'output')
- optional Open the QGIS projects located at data/ to explore the spatial data
- Run the Julia scripts to perform CT computations:
- natural.jl for the natural routes
- campaign.jl for the three sources (Massalia, Tarraco and Emporion)
- local.jl for the local analysis around Puig Ciutat
- outputs will be stored at the folder output; they are already displayed within the QGIS projects.
- data -> data used for the analysis
- base -> basic data used for all analysis. It includes coastline, DEMs and frictions maps, buffer areas, and layers of points used for statistical analysis
- campaign/local/natural.qgz -> qgis projects including all layers for each of the three analysis
- campaign -> landmarks and sampled connectivity values for the three campaign-based scenarios (Massalia, Tarraco and Emporion)
- local -> data required for the local analysis. It includes sampled points, the perimeter of the site and cumulative viewshed analysis
- natural -> vectorial data required for the connectivity analysis of the entire case study area
- qgisColorRamps -> color ramps for rasters and maps; it includes DEM, friction costs, and CT current rasters. The file compute_quantiles_conn.R was used to define the 10/25/50/75/90 thresholds for the different current rasters
- output -> outcome of Circuistcape
- campaign -> *CT current raster for the three campaign-based scenarios (Massalia, Tarraco and Emporion); LCP output
- local -> CT cumulative current raster for the local region used in the third analysis
- natural -> CT cumulative current raster for the entire case study area