Reclassification of raster

Is there a better way to do this?

I'm trying to create a script that will change classifications of a raster to:

0 for: 0,7-9,15-20,40,62-63,73,78-87,89-130,132-151,153-175,177-203,215,228,251-253

1 for: all others

#RECLASSIFY LANDCOVER #Considers 0 for: 63,81-83,87,111-112,121-124,141-143,190,195, #Considers 1 for: All others. #Considers "NODATA" for: background, blank import arcpy from arcpy import env from import * env.workspace = r'C:ErosionLandCover' outReclass1 = Reclassify("CDL_2014.tif", "Value", RemapRange([[0,0,"NODATA"],[7,9,"NODATA"],[15,20,"NODATA"],[40,40,"NODATA"], [62,62,"NODATA"],[63,63,0],[73,73,"NODATA"],[78,80,"NODATA"],[81,83,0], [84,86,"NODATA"],[87,87,0],[89,91,"NODATA"],[93,110,"NODATA"],[111,112,0], [113,120,"NODATA"],[121,124,0],[125,130,"NODATA"],[132,140,"NODATA"], [141,143,0],[144,151,"NODATA"],[153,175,"NODATA"],[177,189,"NODATA"], [190,190,0],[191,194,"NODATA"],[195,195,0],[196,203,"NODATA"], [215,215,"NODATA"],[228,228,"NODATA"],[251,253,"NODATA"]]), 1)'C:ErosionOutputCDL_reclass.tif')

I was attempting to modify the script from the ArcGIS Resource Center Desktop Help.

You should save the output raster to a raster format, not to a geodatabase, e.g.:"C:output

EDIT: And there is a missing bracket in the Reclassify line (RemapRange is enclosed but not Reclassify). I've updated my code as well.

You can also simplify theremapparameter a little bit by using themissing_valuesparameter for all other values that must be remapped to 1:

outReclass1 = Reclassify("raster.tif", "Value", RemapRange([[0,0,0],[7,9,0],[15,20,0], [40,40,0],[62,63,0],[73,73,0],[78,87,0], [89,130,0],[132,151,0],[153,175,0], [177,203,0],[215,215,0],[228,228,0],[251,253,0]]), 1)

See the Reclassify help page for more details and examples.

Reclassify (3D Analyst)

If a range of values is to be reclassed, the ranges should not overlap except at the boundary of two input ranges. Where overlapping occurs, the higher end of the lower input range is inclusive, and the lower end of the higher input range is exclusive.

For example, if two ranges are specified, such as reclassifying values 1 to 5 as 100 and values 5 to 10 as 200, an input value less than or equal to 5 will be assigned the value 100 in the output, and an input value that is larger than 5, such as 5.01, will be assigned to 200.

In the tool dialog, the Classify or Unique options in the Reclassification parameter allows you to generate a remap table based on the values of the input raster. The Classify option opens a dialog and allow you to specify a method from one of the Data classification methods and number of classes. The Unique option will populate the remap table using the unique values from the input dataset.

It is recommended to calculate statistics on a mosaic dataset before reclassifying the data.

From the tool dialog, the remap table can be stored for future use with the Save option. You can save the remap to any relational table format. Use the Load option to reload remap tables you previously created with the Save button.

It is recommended to only load tables previously saved by the Reclassify tool. The table format is specific and must contain the fields FROM , TO , OUT , and MAPPING .

If the input raster has an attribute table, it will be used to create the initial reclassification table. If the input raster does not have an attribute table, you can run the Build Raster Attribute Table tool from the Data Management toolbox to build one before inputting the raster into the Reclassify tool. Otherwise, when you input the raster, a reclassification table will be created for it by first applying geoprocessing environment settings, such as Extent and Cell size, and scanning the raster.

When the input raster is a layer from Contents , the default reclassification table will import the unique values or classified break values as specified by the layer symbology. The current geoprocessing environment settings will be ignored when importing those values. Otherwise, the reclassification must be manually entered or generated using the unique or classification options.

Once the remap table of the reclassification has been modified, the table will not be updated if a new input raster is selected. If the reclassification is not suitable for the new raster, a new reclassification can be reinitialized by one of the following methods

  • Remove all remap records using the erase option and manually add the new values.
  • Select the unique or classification options to generate a new reclassification.

When using the Reclassify tool as part of a model

  • If the input to the tool is derived data from a tool that isn't already run, the remap parameter in the Reclassify tool will be empty until the preceding tool is run and the model is validated. To avoid this, always run preceding tools before connecting their output variables as input to the Reclassify tool. Alternatively, you can create a custom reclassification table by adding entries.
  • If exposing the reclassification table as a model parameter, the reclass field must be exposed as a variable however, it does not need to be set as a model parameter. If the field is not exposed as a variable, the classify and unique values buttons will be disabled in the model tool dialog box.

By default, this tool will take advantage of multicore processors. The maximum number of cores that can be used is four.

To use fewer cores, use the parallelProcessingFactor environment setting.


The input raster to be reclassified.

Field denoting the values that will be reclassified.

A remap list that defines how the values will be reclassified.

The remap list is composed of three components: From, To, and New values. Each row in the remap list is separated by a semicolon, and the three components are separated by spaces. For example

The output reclassified raster.

The output will always be of integer type.

Denotes whether missing values in the reclass table retain their value or get mapped to NoData.

  • DATA —Signifies that if any cell location on the input raster contains a value that is not present or reclassed in a remap table, the value should remain intact and be written for that location to the output raster. This is the default.
  • NODATA —Signifies that if any cell location on the input raster contains a value that is not present or reclassed in a remap table, the value will be reclassed to NoData for that location on the output raster.

Advanced Distributed Runoff Model Calibration and Accuracy

An event-based, kinematic, infiltration-excess, distributed rainfall-runoff model was developed to acknowledge and account for the spatial variability and uncertainty of several parameters relevant to storm surface runoff production and surface flow. The model is compatible with raster Geographic Information Systems (GIS) and spatially and temporally varied rainfall data. Monte Carlo simulation and a likelihood measure are utilized to calibrate the model allowing for a range of possible system responses from the calibrated model. Using rain gauge adjusted radar-rainfall estimates, the model was applied and evaluated to a limited number of historical events for two watersheds within the Denver Urban Drainage and Flood Control District (UDFCD) that contain mixed land use classifications. The 95% uncertainty bounds obtained from the model envelop almost all observed responses at the main basin outlets for the events considered, suggesting an acceptable model structure. While based on a limited number of Monte Carlo simulations and considered events, for the two basins that were considered, Nash arid Sutcliffe efficiency scores of 0.88/0.10, 0.14/0.71, and 0.99/0.95 for runoff volume, peak discharge, and time to peak, respectively, were obtained from the model.

Geometric Correction

Geometric Correction
geometric correction - [remote sensing] The correction of errors in remotely sensed data, such as those caused by satellites or aircraft not staying at a constant altitude or by sensors deviating from the primary focus plane.

You read in Chapter 6 that scale varies in unrectified aerial imagery due to the relief displacement caused by variations in terrain elevation.

3 Geometric Correction
Raw digital images acquired by earth observation systems usually contain geometric distortions so they do not reproduce the image of a grid on the surface faithfully.

model. Choose polynomial as a generic correction tool.

The correction of errors in remotely sensed data, such as those caused by satellites or aircraft not staying at a constant altitude or by sensors deviating from the primary focus plane.

-Image-processing procedure that corrects spatial distortions in an image. geostationary-Refers to satellites traveling at the angular velocity at which the earth rotates as a result, they remain above the same point on earth at all times.

The bands have # # been coregistered both between and within telescopes, and the data have been # # resampled to apply the

s. As for the Level 1A product, these # # Level 1B radiances are generated at 15m, 30m, and 90m resolutions corresponding # # to the VNIR, SWIR, and TIR channels.

of systematic effects (panoramic effect, Earth curvature and rotation). Internal distortions of the image are corrected for measuring distances, angles and surface areas. Specially designed product for photo-interpreting and thematic studies.

3. The calculation of new DN for pixels created during

of a digital scene, based on the values in the local area around the uncorrected pixels.

Heads-up digitizing to capture vector data
Some raster geographic information systems GIS operations such as slope, aspect, and buffer calculations
Import from and export to various standard image file formats such as GeoTIFF.

or registration of an image to a new coordinate system. R2V supports both Bi-linear and Delaunay triangulation methods for geometric transformation.

Materials and methods

Commonly employed ENM approaches include climate matching, basic multivariate logistic regression analyses, and models based on environmental similarity or distance ( Guisan & Theurillat, 2000 James & McCulloch, 2002 Scott et al., 2002 Kadmon et al, 2003 Rushton et al., 2004 ). These techniques are typically deterministic, focussing on a single decision rule or a small set of rules to describe the potential distribution of a particular species. Because distributional limits can be governed by different factors across a species’ geographic range ( Grinnell, 1917 Swihart et al., 2003 ), more complex, multiple-criterion ENM approaches are desirable ( Elith et al., 2006 ). What is more, comparisons among methods are still few, and have focused on interpolation challenges, whereas the challenge at hand in this paper is one of extrapolation to continental and even global scales, so choice of methodology is not clear.

One such multiple-criterion ENM that has seen broad application in particular to extrapolative challenges ( Peterson, 2003 ) is the Genetic Algorithm for Rule-Set Prediction ( garp ), a machine-learning application. Machine-learning algorithms are computationally powerful because they are nonparametric, nonlinear and relatively unaffected by collinearity, and can explore and describe complex relationships among variables in complex solution spaces ( Lek et al., 1996 Olden, 2000 ). In the present study the decision to model the ecological niche of P. ramorum using garp was made because of its multiple decision-rule capability, known robust predictive performance ( Peterson & Cohoon, 1999 Stockwell & Peterson, 2002 Anderson et al., 2003 ), and ability to project niche models onto different geographic areas.

garp is a superset of modelling algorithms (e.g. logistic regression, range rules) that reconstructs ecological niches of species based on nonrandom associations between occurrence points and the ecological characteristics of those localities ( Stockwell & Noble, 1992 Stockwell, 1999 Stockwell & Peters, 1999 ). Such models can be used to predict species’ potential geographic distribution ( Peterson et al., 2002a,b Peterson, 2003 Soberón & Peterson, 2005 ). Model output is spatially explicit, with potential distributions summarized in arc / info raster grids. Detailed descriptions of the garp algorithm, covered briefly below, are presented by Stockwell & Noble (1992 ) and Stockwell & Peters (1999 ).

garp model inputs are species occurrence data (latitude/longitude) and environmental variables. Generally, a minimum of 20 occurrence points are needed to produce models that aren't overfit and unable to predict into unsampled areas ( Stockwell & Peterson, 2002 ). Basic environmental inputs are raster geographic information system (GIS) coverages (see detailed explanations below) consisting of topographic and climate variables biotic variables can also be included (e.g. a coverage representing host density), but typically are not used because these data are often lacking and/or difficult to represent in a spatially explicit manner. garp partitions occurrence data into training and test datasets, the former used to build the model, and the latter used to test the accuracy of the model.

garp describes nonrandom relationships between a species’ occurrence and the environment using multiple rules. Rules are if … then statements relating to environmental conditions and whether a species is present or absent under them. garp employs four rule types: atomic rules (use only single values of a variable), envelope rules (spanning ranges of values), negative range rules (as envelope rules, but with ranges being exluded from the prediction rather than included), and logit rules (logistic regression models) ( Stockwell & Peters, 1999 ). After building an initial set of rules, garp makes iterative changes to them: in each iteration (‘generation’), rules undergo ‘genetic’ changes – insertions, deletions, point mutations, and crossing over among rules can all occur. In this manner, garp explores solution space flexibly to identify nonrandom relationships between environmental conditions and species’ occurrences. The final rule set is arrived at when iterative changes fail to improve the model's predictive accuracy, or when a user-defined number of iterations has been reached.

garp is not a deterministic algorithm, so its output (rule sets and associated predicted distributional areas) varies from run to run. To capture the variation among potential distributions, a best-subset approach can be employed, where the best models, as determined from overprediction (commission) and underprediction (omission) errors, are selected from the final model set ( Anderson et al., 2003 ). These models can then be overlaid in a GIS to determine where they converge, that is, where spatially most or all of the models agree. garp is not a black box – it is possible to examine the rule sets, and a jackknife feature allows correlations to be made between model performance and input variables, providing some insight into how specific variables affect predictive accuracy ( Peterson & Cohoon, 1999 Peterson et al., 2003a ).

Locality records (116) for North American P. ramorum (A2 mating type see Discussion) were acquired from the OakMapper web application ( and the Oregon Department of Agriculture (

Because garp models distributions based on an organism's presence/absence within a landscape pixel (in this analysis, pixels were 0·08 × 0·08° see below), one point was randomly selected from each pixel where P. ramorum occurred, yielding 68 spatially unique localities (Fig. 1a). From these records, a training dataset was compiled by randomly selecting a single point from each 0·25 × 0·25° block (∼28 × ∼28 km), ensuring input of at least 20 datapoints and capturing the overall spatial distribution of P. ramorum records. Based on this stratification, 25 points were used to build the models and 43 points for testing predictive accuracy, with models trained and results mapped at the finer 0·08° resolution.

Predicted potential distribution for Phytophthora ramorum in California counties, based on known occurrence points (shown as yellow squares, inset) and three distinct environmental data sets, generated using the Genetic Algorithm for Rule-set Prediction (model outputs, which ranged 0–10, were simply added to produce this visualization). Models indicate the potential for spread more broadly along the coast, as well as to the western slopes of the Sierra Nevada. Shading denotes the number of models predicting potential presence in a particular pixel.

Raster coverages (again, pixel size 0·08°) were assembled summarizing geographic distributions for 35 species of red oaks north of Mexico, as follows: Quercus acerifolia, Q. agrifolia, Q. arkansana, Q. buckleyi, Q. coccinea, Q. ellipsoidalis, Q. emoryi, Q. falcata, Q. georgiana, Q. graciliformis, Q. gravesii, Q. hemisphaerica, Q. hypoleucoides, Q. ilicifolia, Q. imbricaria, Q. incana, Q. inopina, Q. kelloggii, Q. laevis, Q. laurifolia, Q. marilandica, Q. myrtifolia, Q. nigra, Q. pagoda, Q. palustris, Q. parvula var. shrevii, Q. phellos, Q. pumila, Q. robusta, Q. rubra, Q. shumardii, Q. texana, Q. velutina, Q. viminea ( Nixon, 1980 ) and Q. tardifolia, but excluding Q. wislizeni, as it has not yet been reported as a foliar or bark canker host of P. ramorum ( Dodd et al., 2004 ). Two non-Quercus host trees, L. densiflorus and Arbutus menziesii, were also included. Distributional maps of 24 oaks and the two non-oak hosts were available as shapefiles (, which were converted into raster format. For the 11 remaining oak species, comparable datasets were created based on published range maps and distributional information ( All host distributions were overlaid and aggregated over North America to derive a coverage of potential host tree distribution and used as a landscape mask to define the spatial extent across which SOD might be expected to occur in relation to host tolerances (see below).

Topographic and climatic inputs were from the Hydro-1 K dataset (U.S. Geological Survey USGS) data describing topography (aspect, elevation, flow accumulation, flow direction, topographic index, monthly normalized difference vegetation index (NDVI) from years 1996–2000 derived from advanced very high resolution radiometer (AVHRR) imagery (, and yearly mean values (1961–1990) of diurnal temperature range, ground frost frequency, minimum, maximum and mean temperatures, precipitation, solar radiation, vapour pressure, and wet-day frequency drawn from the Intergovernmental Panel on Climate Change (IPCC) data warehouse ( USGS and NDVI data were 0·01 × 0·01° base resolution (1 × 1 km), generalized to 0·08 × 0·08° (8 × 8 km) for analysis IPCC data were 0·5 × 0·5° resolution (50 × 50 km) subdivided to 0·08 × 0·08° (8 × 8 km) for analysis. Each of the three environmental data sources provides a distinct sort of information: topography summarizes landscape form climatic data inform about the atmospheric conditions above that landscape and NDVI provides a view of seasonal patterns of greenness across landscapes (obviously at least in part a function of topography and climate). NDVI can be used to infer climatic conditions, but tends to correlate well only with minimum temperature, mean temperature and precipitation, and soil moisture ( Wang et al., 2001 Adegoke & Carleton, 2002 ).

Based on these tradeoffs, five distinct ecological niche models were generated, differing in the type and spatial extent of environmental variables: NDVI + topography (constrained within the spatial distribution of potential host trees), NDVI + topography, climate + topography modelled over all of North America, and regional models of the latter two combinations. Regional models were based on environments represented within a 500-km buffer around known occurrence points.

For each set of models, the 10 best replicate garp runs were chosen based on the combination of omission and commission error components calculated from the independent testing data mentioned above ( Anderson et al., 2003 ). These ‘best subsets’ of models were summed to produce maps of potential SOD distribution. In these summary outputs, values ranged from 0 (areas where no models predicted potential presence) to 10 (areas where all models predicted potential presence). In general, areas in which all best-subsets models agreed in predicting potential presence were focussed on. The best-subset models were then projected to other regions to outline potential geographic distributions for the pathogen in the USA and worldwide ( Peterson & Vieglais, 2001 Peterson, 2003 ). This approach has seen extensive testing in a variety of invasive species systems, and has shown excellent abilities in predicting the geographic course of already-established invasions ( Peterson & Vieglais, 2001 Papes & Peterson, 2003 Peterson, 2003 Peterson et al., 2003a,b ).

To evaluate the accuracy of the garp predictions, the area under the receiver operating characteristic ( roc ) curve of the model's predictive performance was determined. An roc curve represents the relationship between the probabilities of a model correctly classifying true positives (sensitivity Y-axis) and incorrectly classifying false positives (1 – specificity X-axis) ( Fielding & Bell, 1997 ). If these probabilities are equal, the model has no discriminatory power and will perform no better than random this relationship is represented as a ‘line of no information’, a 1:1 relationship between X and Y axes (slope = 1) under which the area is 0·5 ( Hanley & McNeil, 1982 ). As such, the area under curve ( auc ) is the probability at which the model will make a successful classification, ranging from random (0·5) to perfect discrimination (1·0). auc values of 0·81–1·00 are considered robust ( Swets, 1988 ).

garp auc was calculated using a Wilcoxon test ( Hanley & McNeil, 1982 ) to compare the number of landscape pixels where 0, 1, 2, etc., best-subset models predicted P. ramorum potential presence vs. the number of those pixels where the phytopathogen actually occurred (testing data n = 25). The statistical significance of each model set was determined with a Z-test of the garp auc versus that of a random model (0·5 Hanley & McNeil, 1982 ).

Reclassification of raster - Geographic Information Systems


This paper describes a map algebra language that can form the foundation for overlay and neighborhood analyses in both raster geographic information systems and image processing systems. The language combines maps and images in expressions with arithmetic and logical operators and mathematical functions to produce new maps and images. This algebra has been implemented in the GRASS r.mapcalc module. The language syntax is first described and then examples relevant to GIS and image processing applications are presented. 1 Introduction 1.1 Background Geographic information systems (GIS) have been evolving from basic spatial data storage, retrieval, and display systems to more powerful spatial modeling systems. This transition is mainly due to more robust query capabilities. The creation of spatial query languages that empower the end-user to design and create spatial models will advance this development even further. Tomlin (1990) describes a spatial modeling language for raster data ..

To submit an update or takedown request for this paper, please submit an Update/Correction/Removal Request.

Suggested articles

Useful links

Writing about CORE?

Discover our research outputs and cite our work.

CORE is a not-for-profit service delivered by the Open University and Jisc.

1. Introduction

About 85 million people are without access to electricity in Ethiopia, majority of them are living rural parts of the country (Michael, 2018 ). Electricity access is a key for sustainable development of any country. However, in developing countries like Ethiopia, it is difficult to electricity a sparsely populated community. Lack of scientific and methodological know-how as regards planning, site selection, distribution, and density of population settlement, economic level, social levels, and distance from the national grid are main factors to lowest electrification rates in Ethiopia. Thus, this study introduces renewable energy systems as a possible alternative for rural electrification (Qerimi et al., 2020 ).

Ethiopia is endowed with so many renewable energy potentials. As per the study, the estimated potential for hydropower is 45GW, for wind is 10GW, for geothermal 5GW, and the potential for solar irradiation ranges from 4.5kwh/m2/day to 7.5kwh/m2/day (Mondal et al., 2018 ) (Falchetta et al., 2019 ). With all the resources potential available, improper electrification planning is a challenge in Ethiopia to reach the required level of standard in rural electrification. Ethiopian national electrification program implementation roadmap presents action plan for achieving nationwide universal electricity access by 2025 (Korkovelos et al., 2019 ) (Drouin, 2018 ). Fast-paced ambitious grid connections rollout program to increase access by 65% of the population without access is one of the programs priorities. The other option established in the action plane is the enhanced design and reach of grid access rollout program to provide the remaining 35% of the population resided in the rural and deep rural households by 2025 using solar home systems and isolated mini/micro grids (Mondal et al., 2018 ). However, there is no appropriate electrification planning to meet this target.

A Geographic information system (GIS) based graphics management system is important for rural electrification planning. The integration of basic information and geographic information enables rural electrification planning easy. Basic information and geographic information are considered in developing a geospatial information system to obtain the best possible electrification planning and strategies. Data like sunshine hour/irradiance, wind speed, slope, land use-land cover, protected areas, water bodies, forest, and towns are collected and analyzed. These data are used to identify the optimal electrification options for particular areas in four districts of the Zone (Fogera, Dera, Farta, and Este). The weighting factors of this criterion are also determined using Analytic Hierarchy Process (AHP) (Alami Merrouni et al., 2018 ). These weighted and reclassified values are multiplied to produce the final map of electrification options.

There is different energy planning system in which geographical information system-related studies in which multi-criteria decision making and geographic information system have been used for different perspectives. Geographic information system has been used for energy planning. The existing literature review explored that different researchers used different energy planning methods. This study contributes to the existing literature by proposing more available energy resource options nearby the community.

Besides integrating renewable energy resources like solar, wind, hydropower, geothermal, and fossil, Ethiopia is developing large-scale hydropower to achieve universal electrification program. However, the challenges of the preferred option and mix in cost effective way based on geospatial procedure to determine accurately is not sounded to achieve the goal. Therefore, the objective of this study is to study the introduction of renewable energy systems as a possible alternative for rural electrification by using geospatial technologies in South Gondar Zone.


  • Riegelmann, Edward A.
  • The Cape Verde Islands, off the coast of West Africa, are subject to violent precipitation events that cause extensive soil erosion and degradation of the landscape. A systematic approach to the collection, analysis, and dissemination of hydrophysical and geographic data was needed to effectively analyze the cause and effect variables of landscape erosion at the watershed scale. A methodology incorporating geographic information system (GIS) analysis and modified stream reach inventory and channel stability evaluation techniques was developed for Ribeira da Boca Larga drainage on Santiago Island, Republic of Cape Verde. The stream reach inventory employed is a modified version of Pfankuch's "Stream Reach Inventory and Channel Stability Evaluation" (1975). Modifications were made due to the unique environmental conditions encountered in Cape Verde. The geographic information system runs on an IBM PC-AT compatible micro-computer and consists of three popular vector and raster geographic data storage and analysis software packages linked together by a fourth software bundle developed for custom communication and analysis of geographic data. Data from the stream reach inventory and geographic information system were combined to create uniform channel reach analyses and automated GIS products. These products are designed to assist the watershed manager, land use planner, and agricultural engineer in developing a concerted response to erosional phenomena in the evaluated drainage. The integrated stream reach inventory and geographic information system methodology was developed to serve as a model for watershed analysis throughout the Cape Verde Archipelago.
  • Oregon State University
  • File scanned at 300 ppi (Monochrome) using ScandAll PRO 1.8.1 on a Fi-6770A in PDF format. CVista PdfCompressor 4.0 was used for pdf compression and textual OCR.
  • description.provenance : Approved for entry into archive by Patricia Black([email protected]) on 2012-07-19T20:28:56Z (GMT) No. of bitstreams: 1RiegelmannEdwardA.pdf: 52135806 bytes, checksum: b4a6b13770c6b27ba78a0c955af3269a (MD5)
  • description.provenance : Submitted by Kaylee Patterson ([email protected]) on 2012-07-18T21:33:48ZNo. of bitstreams: 1RiegelmannEdwardA.pdf: 52135806 bytes, checksum: b4a6b13770c6b27ba78a0c955af3269a (MD5)
  • description.provenance : Made available in DSpace on 2012-07-19T20:28:56Z (GMT). No. of bitstreams: 1RiegelmannEdwardA.pdf: 52135806 bytes, checksum: b4a6b13770c6b27ba78a0c955af3269a (MD5) Previous issue date: 1989-02


Bivand, Roger S., Edzer Pebesma, and Virgilio Gómez-Rubio. 2013. Applied Spatial Data Analysis with R. 2nd ed. 2013 edition. New York: Springer.

Chambers, John M. 2016. Extending R. CRC Press.

Hijmans, Robert J. 2017. Raster: Geographic Data Analysis and Modeling.

Longley, Paul, Michael Goodchild, David Maguire, and David Rhind. 2015. Geographic Information Science & Systems. Fourth edition. Hoboken, NJ: Wiley.

Pebesma, Edzer. 2018. “Simple Features for R: Standardized Support for Spatial Vector Data.” The R Journal.

[1] Note also that RPyGeo provides access to ArcMap which is a commercial Desktop GIS software.