Package 'GeoFIS'

Title: Spatial Data Processing for Decision Making
Description: Methods for processing spatial data for decision-making. This package is an R implementation of methods provided by the open source software GeoFIS <https://www.geofis.org> (Leroux et al. 2018) <doi:10.3390/agriculture8060073>. The main functionalities are the management zone delineation (Pedroso et al. 2010) <doi:10.1016/j.compag.2009.10.007> and data aggregation (Mora-Herrera et al. 2020) <doi:10.1016/j.compag.2020.105624>.
Authors: Serge Guillaume [aut], Jean-Luc Lablée [aut, cre], INRAE [cph] (National Research Institute for Agriculture, Food and Environment, France)
Maintainer: Jean-Luc Lablée <[email protected]>
License: CeCILL
Version: 1.1.0
Built: 2024-11-03 06:32:38 UTC
Source: CRAN

Help Index

Class "AggregFis"

Description

The Fis aggregation operator to be used in Fusion

Slots

fis

FisPro::Fis object, The Fis to be used in the aggregation operator

output_index

integer value, The index (1-based index) of the output in the Fis to be used in the aggregation

See Also

NewAggregFis

Aggregation using linguistic rules

Class "AggregFunction"

Description

The functional aggregation operator to be used in Fusion

Slots

func

Function, The function used for the aggregation

See Also

NewAggregFunction

Class "AggregOwa"

Description

The OWA aggregation operator to be used in Fusion

Slots

weights

numeric vector, The weights of the OWA aggregation operator (the sum of the weights must be equal to 1 without negative values)

See Also

NewAggregOwa

Aggregation using numerical operators

Class "AggregWam"

Description

The WAM aggregation operator to be used in Fusion

Slots

weights

numeric vector, The weights of the WAM aggregation operator (the sum of the weights must be equal to 1 without negative values)

See Also

NewAggregWam

Aggregation using numerical operators

Soil conductivity 2014 dataset

Description

The soil conductivity of a vine plot in year 2014

Usage

data(conductivity_2014)

Format

sp::SpatialPointsDataFrame object with 353 observations and 1 attribute:

conduct

numeric value, The soil conductivity

Border dataset

Description

The soil conductivity border of a vine plot

Usage

data(conductivity_border)

Format

sp::SpatialPolygonsDataFrame object with 1 polygon delimiting the border of the vine plot:

id

integer value, The id of the polygon

Classify data in management zones

Description

Classify data in management zones of maps obtained with the Zoning algorithms.

Usage

DataInZone(data, maps, use_id = FALSE)

Arguments

data

sp::SpatialPointsDataFrame or sp::SpatialMultiPointsDataFrame object, The input data.
maps

sp::SpatialPolygonsDataFrame object, or a list of sp::SpatialPolygonsDataFrame, The map or list of maps to process.
use_id

boolean, Use the id attribute of the zone in the map if TRUE, or the order index (1-based indexed) of the zone in the map if FALSE (the default).

Value

sp::SpatialPointsDataFrame or sp::SpatialMultiPointsDataFrame object depending of the type of input data. Return table of membership of each point in the zones for each map, the input data and a column for each map processed (named "map{number of zones in the map}") with the class of each point.

The "Euclidean" distance

Description

Function to create an "Euclidean" distance
To be used with the Zoning combine_distance or attribute_distance field

Usage

EuclideanDistance()

Value

Euclidean distance object

Class "Fusion"

Description

The main class to perform data fusion
More information is available in the vignette "Data Fusion with GeoFIS"

Active bindings

aggregate

data.tree::Node object, or a list of data.tree::Node, The node(s) to aggregate

Methods

Public methods

Method new()

The constructor to build an object of class Fusion.

Usage
Fusion$new(source)
Arguments
source

data.frame or sp::Spatial*DataFrame object of sp::sp package
Keep only numeric attributes

Method perform()

Perform the data fusion

Usage
Fusion$perform()

Method output()

Get the output aggregated data (same object type as data source)

Usage
Fusion$output()
Returns

data.frame or sp::Spatial*DataFrame object

References

Guillaume S, Bates T, Lablee J, Betts T, Taylor J (2020). “Combining Spatial Data Layers Using Fuzzy Inference Systems: Application to an Agronomic Case Study.” In Proceedings of the 6th International Conference on Geographical Information Systems Theory, Applications and Management (GISTAM 2020), 62-71. ISBN 978-989-758-425-1.

Mora-Herrera DY, Guillaume S, Snoeck D, Zuniga Escobar O (2020). “A fuzzy logic based soil chemical quality index for cacao.” Computers and Electronics in Agriculture, 177, 105624. doi:10.1016/j.compag.2020.105624.

See Also

NewFusion

Data Fusion documentation

Examples

# more information about this example in the vignette "Data Fusion with GeoFIS"
# section "Learning illustration"

library(GeoFIS)

data(fusion_cars)

fusion <- NewFusion(fusion_cars)
a <- NewFusionInput("a", NewMfTrapezoidalInf(4, 20), "A")
v <- NewFusionInput("v", NewMfTrapezoidalSup(100, 500), "V")
s <- NewFusionInput("s", NewMfTrapezoidalSup(120, 220), "S")
c <- NewFusionInput("c", NewMfTrapezoidalInf(6, 16), "C")
owa_aggreg <- NewFusionAggreg("score", NewAggregOwa(c(1, 0, 0, 0)), a, v, s, c)
fusion$aggregate <- owa_aggreg
fusion$perform()
score <- fusion$output()["score"]
print(score)

Fusion Cars dataset

Description

Illustration dataset for data fusion numerical operators learning

Usage

data(fusion_cars)

Format

data.frame object with four cars described by four attributes:

A

numeric value, the acceleration time (s) from 0 to 100 km/h

V

numeric value, the volume of the trunk (l)

S

numeric value, the maximum speed (km/h)

C

numeric value, the gas consumption (l per 100 km)

Class "FusionLabel"

Description

Defines the allowed labels for the FisPro::Mfs of the fuzzy inputs or output in the FisPro::Fis "Fusion"

Active bindings

very_low

character vector (read-only), The very_low label

low

character vector (read-only), The low label

average

character vector (read-only), The average label

high

character vector (read-only), The high label

very_high

character vector (read-only), The very_high label

Methods

Public methods

Method get_labels()

Get the allowed labels depending on the granularity in the FisPro::Fis
for granularity 2, allowed labels are: [low, high]
for granularity 3, allowed labels are: [low, average, high]
for granularity 4, allowed labels are: [very_low, low, high, very_high]
for granularity 5, allowed labels are: [very_low, low, average, high, very_high]

Usage
FusionLabel$get_labels(granularity)
Arguments
granularity

integer value, The granularity of the fuzzy inputs or output in the FisPro::Fis (value in range [2, 5])

Returns

character vector, The allowed labels for the granularity

The "Fuzzy" distance

Description

Function to create a "Fuzzy" distance
The fuzzy distance function is based on a fuzzy partition that allows for integrating expert knowledge into distance calculations
To be used with the Zoning attribute_distance field

Usage

FuzzyDistance(fisin)

Arguments

fisin

FisPro::FisIn object, The partition used for the fuzzy distance (must be a standardized fuzzy partition)

Value

Fuzzy distance object

References

Guillaume S, Charnomordic B, Loisel P (2013). “Fuzzy partitions: a way to integrate expert knowledge into distance calculations.” International Journal of Information Sciences, 245, 76-95. doi:10.1016/j.ins.2012.07.045.

Guillaume S, Charnomordic B (2013). “Fuzzy partition-based distance practical use and implementation.” In CFP12FUZ-USB ICN (ed.), IEEE International Conference on Fuzzy Systems, paper F-1136.

Learn the OWA weights

Description

Learn the OWA weights using a non-negative least-square optimization method with the constraint that the sum of weights must be equal to 1. The input values are previously sorted in increasing order. The resulting weights are given from min to max. More information is available in the vignette "Data Fusion with GeoFIS", section "Learning illustration".

Usage

LearnOwaWeights(data, target, digits = 3)

Arguments

data

data.frame or numeric matrix, The input data (all columns must be in range [0, 1])
target

numeric vector, The target data (must be in range [0, 1])
digits

integer value, The number of digits to which weights are to be rounded (default is 3)

Value

numeric vector, The OWA weights

Learn the WAM weights

Description

Learn the WAM weights using a non-negative least-square optimization method with the constraint that the sum of weights must be equal to 1.
More information is available in the vignette "Data Fusion with GeoFIS", section "Learning illustration".

Usage

LearnWamWeights(data, target, digits = 3)

Arguments

data

data.frame or numeric matrix, The input data (all columns must be in range [0, 1])
target

numeric vector, The target data (must be in range [0, 1])
digits

integer value, The number of digits to which weights are to be rounded (default is 3)

Value

numeric vector, The WAM weights

The "Maximum" distance

Description

Function to create a "Maximum" distance
To be used with the Zoning zone_distance field

Usage

MaximumDistance()

Value

Maximum distance object

The "Mean" distance

Description

Function to create a "Mean" distance
To be used with the Zoning zone_distance field

Usage

MeanDistance()

Value

Mean distance object

The "Minimum" distance

Description

Function to create a "Minimum" distance
To be used with the Zoning zone_distance field

Usage

MinimumDistance()

Value

Minimum distance object

The "Minkowski" distance

Description

Function to create a "Minkowski" distance
To be used with the Zoning combine_distance field

Usage

MinkowskiDistance(power = 2)

Arguments

power

numeric value, The power of the Minkowski distance
The default value is 2 (equivalent to euclidean distance)

Value

Minkowski distance object

Create object of class "AggregFis"

Description

Function to create an aggregation operator of class AggregFis to be used in Fusion

Usage

NewAggregFis(fis, output_index = 1)

Arguments

fis

FisPro::Fis object, The Fis to be used in the aggregation operator
output_index

integer value, The index (1-based index) of the output in the Fis to be used in the aggregation (the default is 1)

Value

AggregFis object

See Also

Aggregation using linguistic rules

Create object of class "AggregFunction"

Description

Function to create an aggregation operator of class AggregFunction to be used in Fusion

Usage

NewAggregFunction(func)

Arguments

func

The function to be used for the aggregation

Create object of class "AggregOwa"

Description

Function to create an aggregation operator of class AggregOwa to be used in Fusion

Usage

NewAggregOwa(weights)

Arguments

weights

numeric vector, The weights of the OWA aggregation operator (the sum of the weights must be equal to 1 without negative values)

See Also

Aggregation using numerical operators

Create object of class "AggregWam"

Description

Function to create an aggregation operator of class AggregWam to be used in Fusion

Usage

NewAggregWam(weights)

Arguments

weights

numeric vector, The weights of the WAM aggregation operator (the sum of the weights must be equal to 1 without negative values)

See Also

Aggregation using numerical operators

Create object of class "Fis" to be used in data fusion

Description

Function to create object of class FisPro::Fis to be used in AggregFis

Usage

NewFisFusion(
  fis_name,
  input_names,
  input_granularities,
  output_name,
  output_conclusions
)

Arguments

fis_name

character vector, The name of the Fis
input_names

character vector, The Fis inputs names
input_granularities

integer vector, The granularity (number of membership functions) for each Fis input (granularity must be in range [2, 5])
output_name

character vector, The name of the Fis output
output_conclusions

numeric or character vector, The conclusions of the rules in the Fis
the rules are generated according to the granularity of each input, in the lexicographic order of inputs Mfs
(prod(input_granularities) rules are generated)
if numeric vector, a crisp output FisPro::FisOutCrisp will be added to the Fis (all output conclusions must be be in range [0, 1])
if character vector, a fuzzy output FisPro::FisOutFuzzy will be added to the Fis, the output_conclusions contains the labels of Mfs in the fuzzy output (labels defined on FusionLabel)
the length of output_conclusions must be equal to the number of generated rules.

Value

FisPro::Fis object

See Also

Aggregation using linguistic rules

Create object of class "Fusion"

Description

Function to create object of class Fusion

Usage

NewFusion(...)

Arguments

...

arguments of Fusion constructor

Value

Fusion object

Create an aggregation node to be used in data fusion

Description

Function to create an aggregation node to be used in Fusion

Usage

NewFusionAggreg(name, aggreg, ...)

Arguments

name

character vector, The name of the node
aggreg

Aggreg object, The aggregation operator to be used to compute the aggregation of satisfaction degrees
must be an AggregWam, AggregOwa, AggregFis or AggregFunction object
...

data.tree::Node objects, The nodes to aggregate
can be an input node built with NewFusionInput or an aggregate node built with NewFusionAggreg for a hierarchical aggregation structure

Value

data.tree::Node object

See Also

Aggregation of the degrees

Create an input node to be used in data fusion

Description

Function to create an input node to be used in Fusion

Usage

NewFusionInput(name, mf, attribute = name)

Arguments

name

character vector, The name of the node
mf

FisPro::Mf object, The membership function to be used to compute the satisfaction degree of the input
attribute

character vector, The attribute name in the source dataset (default is the same as name)

Value

data.tree::Node object

See Also

From raw data to satisfaction degrees

Create object of class "Zoning"

Description

Function to create object of class Zoning

Usage

NewZoning(...)

Arguments

...

arguments of Zoning constructor

Value

Zoning object

Tolima dataset

Description

Soil experimental data in three municipalities of Tolima department in Colombia (Mora-Herrera et al. 2020)

Usage

data(tolima)

Format

data.frame object with 30 observations and 8 attributes:

Cadmium

numeric value, Cadmium in Soil (ppm)

pH

numeric value, pH Soil (°pH)

OM

numeric value, Organic Matter (%)

P

numeric value, Available Phosphorus (ppm)

K

numeric value, Exchangeable Potassium (meq/100 g)

BalanceGap

numeric value, Balance Gap (%)

Ngap_N_OpN

numeric value, N Gap (N/Ntarget)

Base_S

numeric value, Base Saturation (%)

References

Mora-Herrera DY, Guillaume S, Snoeck D, Zúñiga Escobar O (2020). “A fuzzy logic based soil chemical quality index for cacao.” Computers and Electronics in Agriculture, 177, 105624. doi:10.1016/j.compag.2020.105624.

The "Area" smallest zone

Description

Function to create an "Area" smallest zone
To be used with the Zoning smallest_zone field

Usage

ZoneArea(area)

Arguments

area

numeric value, The minimum area of the zone to retain the zone in the Zoning process

Value

Area Smallest zone object

The "Size" smallest zone

Description

Function to create a "Size" smallest zone
To be used with the Zoning smallest_zone field

Usage

ZoneSize(number_of_points)

Arguments

number_of_points

integer value, The minimum number of points in the zone to retain the zone in the Zoning process

Value

Size Smallest zone object

Class "Zoning"

Description

The main class to perform zoning
A complete use-case example is described in the vignette "Zoning with GeoFIS"

Active bindings

border

sp::SpatialPolygons object, The border used to limit the processed area, or NULL if the Convex Hull of data source is used
Only data points within the border polygon are processed
The default value is NULL

neighborhood

numeric value, The minimum edge length shared by two Voronoi polygons for being considered as neighbors
or NULL if all contiguous Voronoi polygons are considered as neighbors
The default value is NULL

attribute_distance

list of Distance object (write-only), The functions used to compute the distance between two data points in the attribute space
The length of the list must be equal to the number of zonable attributes, the distance objects are treated in the order of zonable attributes
In case of a single attribute into the zonable dataset, the list is optional and a single Distance object can be provided
Allowed distance objects: EuclideanDistance, FuzzyDistance or NULL if the attribute should not be used in the zoning process
The default value is a list of EuclideanDistance
See Zoning documentation main parameters univariate distance

combine_distance

Distance object (write-only), The function used to combine attribute distances in case of multivariate zoning
Allowed distance objects: EuclideanDistance or MinkowskiDistance
The default value is EuclideanDistance See Zoning documentation main parameters multivariate combination

zone_distance

Distance object (write-only), The function used to compute the distance between 2 zones
Allowed distance objects: MaximumDistance, MinimumDistance or MeanDistance
The default value is MaximumDistance
The pair of zones to be merged are those for which the zone_distance is minimum.
See Zoning documentation main parameters between zone distance

smallest_zone

Smallest zone object (write-only), This criterion is used to determine the smallest size for a zone (number of points or area) to be kept in the final map
Allowed Smallest zone objects: ZoneSize or ZoneArea
The default value is ZoneSize with 1 point

Methods

Public methods

Method new()

Constructor, create a new instance of Zoning

Usage
Zoning$new(source, warn = TRUE)
Arguments
source

sp::SpatialPointsDataFrame or sp::SpatialMultiPointsDataFrame object, The data source

warn

logical value, Show warnings if TRUE, default value is TRUE

Method zonable_data()

Get the zonable data
Keep only the attributes that can be used in the zoning process, meaning numeric atributes, without missing values and with a range that is not limited to a unique value
The last condition is required by the min-max standardization process

Usage
Zoning$zonable_data()
Returns

sp::SpatialPointsDataFrame object

Method perform_voronoi()

Compute the Voronoi diagram

Usage
Zoning$perform_voronoi()

Method voronoi_map()

Get the Voronoi map

Usage
Zoning$voronoi_map()
Returns

sp::SpatialPolygons object

Method perform_neighborhood()

Identify adjacent polygons in the voronoi tesselation

Usage
Zoning$perform_neighborhood()

Method neighborhood_map()

Get the neighborhood map

Usage
Zoning$neighborhood_map()
Returns

sp::SpatialLinesDataFrame object

Method perform_zoning()

Perform the zoning

Usage
Zoning$perform_zoning()

Method map_size()

Get the number of maps with different number of zones available after perform zoning

Usage
Zoning$map_size()
Returns

integer value

Method map()

Get the map corresponding to a number of zones

Usage
Zoning$map(number_of_zones)
Arguments
number_of_zones

integer value, The number of zones in the map

Returns

sp::SpatialPolygonsDataFrame object

Method maps()

Get the maps corresponding to a number of zones

Usage
Zoning$maps(number_of_zones)
Arguments
number_of_zones

integer vector, The number of zones in each map

Returns

list of sp::SpatialPolygonsDataFrame object

References

Pedroso M, Taylor J, Tisseyre B, Charnomordic B, Guillaume S (2010). “A segmentation algorithm for the delineation of management zones.” Computer and Electronics in Agriculture, 70(1), 199-208. doi:10.1016/j.compag.2009.10.007.

Guillaume S, Charnomordic B, Loisel P (2013). “Fuzzy partitions: a way to integrate expert knowledge into distance calculations.” International Journal of Information Sciences, 245, 76-95. doi:10.1016/j.ins.2012.07.045.

Guillaume S, Charnomordic B (2013). “Fuzzy partition-based distance practical use and implementation.” In CFP12FUZ-USB ICN (ed.), IEEE International Conference on Fuzzy Systems, paper F-1136.

See Also

NewZoning

Zoning documentation