Scholarly record
R AND PYTHON BENCHMARKING FOR GEOGRAPHICAL APPLICATIONS
Abstract
Two commonly used programming languages in geosciences are R and Python. Both languages come with inherent advantages and disadvantages. While Python is a general-purpose language with a readable syntax, R is built by statisticians and encompasses their specific language. In contrast with more nominative and static programming languages, like C++ or Java, both R and Python are heavily used in non-informatics disciplines, due to their relative ease of use. Notwithstanding the lower performance of both languages compared to the aforementioned languages, they allow fast prototyping and implementation of algorithms in research contexts. However, depending on the aim of the research, (geo-)scientists will frequently decide between using Python, favoring replicability and accessibility, or R, containing fast-programmable cutting-edge reporting tools. One of the main differences between both languages is that, for Python, with five libraries one can complete most research (Numpy, Pandas, Scipy, Scikit-learn, and Matplotlib), while for R, one should find the right package in the +12.000 available packages on CRAN. The availability of the many packages for R is at the same time one of its greatest strengths. In this research, three common analyses in geosciences are performed in both R and Python. The first analysis highlights the unsupervised classification of (Sentinel-2) satellite imagery. The second series of analysis concerns the transformation of planimetric coordinates from one coordinate reference system to another system. As a final analysis, a series of standard topographic parameters are derived for an equidistant digital elevation model extracted from ASTER- and SRTM- data (EU-DEM). Next to the quantitative benchmarking of the processing time for each analysis, the output of each calculation is investigated as a qualitative benchmark, with a focus on the comprehensibility of the numerical/character output and the visualizations.
Publication Impact Profile
- Captures
- Mendeley - Readers: 12
Publication details
References13
F. Pedregosa et al., “Scikit-learn: Machine Learning in Python,” J. Mach. Learn. Res., vol. 12, pp. 2825–2830, 2011.
H. Wong, C. JA Hartigan, . Series, and undefined 1979, “Algorithm AS 136: A k-means clustering algorithm,” J. R. Stat. Soc. Ser. C (Applied Stat., vol. 28, no. 1, pp. 100–108, 1979.
F. Warmerdam, “The Geospatial Data Abstraction Library,” in Open Source Approaches in Spatial Data Handling, Springer Berlin Heidelberg, 2008, pp. 87–104.
A. Graser and V. Olaya, “Processing: A Python Framework for the Seamless Integration of Geoprocessing Tools in QGIS,” ISPRS Int. J. Geo-Information, vol. 4, no. 4, pp. 2219–2245, Oct. 2015, DOI: 10.3390/ijgi4042219.
J.-P. Donnay and P. Lambot, “BEREF and Lambert 2008 New geodetic and cartographical standards applied in Belgium,” 2012.
T. Knudsen, K. Evers, G. Valsson, and P. Häkli, “The Bricks and Mortar for Contemporary Reimplementation of Legacy Nordic Transformations,” Geophysica, vol. 54, no. 1, pp. 107–108, 2019.
PROJ contributors, “PROJ coordinate transformation software library,” Open Source Geospatial Foundation, 2020. [Online]. Available: https://proj.org/. [Accessed: 16-Jul-2020].
S. Riley, S. DeGloria, and R. Elliot, “Index that quantifies topographic heterogeneity,” Intermt. J. Sci., vol. 5, no. 1–4, pp. 23–27, 1999.
A. Guisan, S. B. Weiss, and A. D. Weiss, “GLM versus CCA spatial modeling of plant species distribution,” Plant Ecol., vol. 143, no. 1, pp. 107–122, 1999, DOI: 10.1023/A:1009841519580.
R. Stone and J. Dugundji, “A study of microrelief—its mapping, classification, and quantification by means of a Fourier analysis,” Eng. Geol., vol. 1, no. 2, pp. 89–187, 1965.
R. Hobson, “Surface roughness in topography: quantitative approach,” Spat. Anal. Geomorphol., pp. 221–245, 1972.
M. F. J. Wilson, B. O’Connell, C. Brown, J. C. Guinan, and A. J. Grehan, “Multiscale terrain analysis of multibeam bathymetry data for habitat mapping on the continental slope,” Mar. Geod., vol. 30, no. 1–2, pp. 3–35, Jan. 2007, DOI: 10.1080/01490410701295962.
D. J. J. Walvoort, D. J. Brus, and J. J. de Gruijter, “An R package for spatial coverage sampling and random sampling from compact geographical strata by k-means,” Comput. Geosci., vol. 36, no. 10, pp. 1261–1267, Oct. 2010, DOI: 10.1016/J.CAGEO.2010.04.005.
View or Download full articleAccess options
SWS access login
Login as SWS Scientific CommitteeLogin as SWS Scientific PartnerLogin as SWS AuthorAuthors and approved SWS contributors will read and export their own linked papers after identity matching by SWS profile, email and SGEM GlobalID.
For librarian assistance: [email protected]
Purchase Instant Access
- Article can be downloaded after successful payment.
- Article may be used according to SWS library access terms.
- Article cannot be redistributed.