Soil organic carbon content in a south area (33 x 33km) of lake Alaotra in Madagascar.

Description

A dataset containing the mean of soil OC content from 0-30 cm layer.

Usage

data(OC)

Format

a RasterLayer with dimensions : 134, 135, 18090 (nrow, ncol, ncell), resolution : 250, 250 (x, y).

Source

ISRIC soilgrid information. HENGL, T., MENDES DE JESUS, J., HEUVELINK, G. B. M., RUIPEREZ GONZALEZ, M., KILIBARDA, M., BLAGOTIC, A., SHANGGUAN, W., WRIGHT, M. N., GENG, X., BAUER-MARSCHALLINGER, B., GUEVARA, M. A., VARGAS, R., MACMILLAN, R. A., BATJES, N. H., LEENAARS, J. G. B., RIBEIRO, E., WHEELER, I., MANTEL, S. & KEMPEN, B. 2017. SoilGrids250m: Global gridded soil information based on machine learning. PLOS ONE, 12, e0169748.

Standard deviation of soil organic carbon content in a south area (33 x 33km) of lake Alaotra in Madagascar.

Description

A dataset containing the standard deviation of soil OC content from 0-30 cm layer.

Usage

data(OC_sd)

Format

a RasterLayer with dimensions : 134, 135, 18090 (nrow, ncol, ncell), resolution : 250, 250 (x, y).

Source

ISRIC soilgrid information. HENGL, T., MENDES DE JESUS, J., HEUVELINK, G. B. M., RUIPEREZ GONZALEZ, M., KILIBARDA, M., BLAGOTIC, A., SHANGGUAN, W., WRIGHT, M. N., GENG, X., BAUER-MARSCHALLINGER, B., GUEVARA, M. A., VARGAS, R., MACMILLAN, R. A., BATJES, N. H., LEENAARS, J. G. B., RIBEIRO, E., WHEELER, I., MANTEL, S. & KEMPEN, B. 2017. SoilGrids250m: Global gridded soil information based on machine learning. PLOS ONE, 12, e0169748.

Soil total nitrogen content in a south area (33 x 33km) of lake Alaotra in Madagascar.

Description

A dataset containing the mean of soil TN content from 0-30 cm layer.

Usage

data(TN)

Format

a RasterLayer with dimensions : 134, 135, 18090 (nrow, ncol, ncell), resolution : 250, 250 (x, y).

Source

ISRIC soilgrid information. HENGL, T., MENDES DE JESUS, J., HEUVELINK, G. B. M., RUIPEREZ GONZALEZ, M., KILIBARDA, M., BLAGOTIC, A., SHANGGUAN, W., WRIGHT, M. N., GENG, X., BAUER-MARSCHALLINGER, B., GUEVARA, M. A., VARGAS, R., MACMILLAN, R. A., BATJES, N. H., LEENAARS, J. G. B., RIBEIRO, E., WHEELER, I., MANTEL, S. & KEMPEN, B. 2017. SoilGrids250m: Global gridded soil information based on machine learning. PLOS ONE, 12, e0169748.

Standard deviation of soil total nitrogen content in a south area (33 x 33km) of lake Alaotra in Madagascar.

Description

A dataset containing the standard deviation of soil TN content from 0-30 cm layer.

Usage

data(TN_sd)

Format

a RasterLayer with dimensions : 134, 135, 18090 (nrow, ncol, ncell), resolution : 250, 250 (x, y).

Source

ISRIC soilgrid information. HENGL, T., MENDES DE JESUS, J., HEUVELINK, G. B. M., RUIPEREZ GONZALEZ, M., KILIBARDA, M., BLAGOTIC, A., SHANGGUAN, W., WRIGHT, M. N., GENG, X., BAUER-MARSCHALLINGER, B., GUEVARA, M. A., VARGAS, R., MACMILLAN, R. A., BATJES, N. H., LEENAARS, J. G. B., RIBEIRO, E., WHEELER, I., MANTEL, S. & KEMPEN, B. 2017. SoilGrids250m: Global gridded soil information based on machine learning. PLOS ONE, 12, e0169748.

Simple check if distribution provided in defineUM() belongs to a list of supported distributions.

Description

Simple check if distribution provided in defineUM() belongs to a list of supported distributions.

Usage

check_distribution(object)

Arguments

Value

TRUE or FALSE.

Author(s)

Kasia Sawicka

Simple check if class of provided object is Spatial

Description

Simple check if class of provided object is Spatial

Usage

check_if_Spatial(object)

Arguments

Value

TRUE or FALSE.

Author(s)

Kasia Sawicka

Converting a spatial correlogram model to a variogram model

Description

Used internally in genSample() in case of sampling by unconditional gaussian simulation.

Usage

crm2vgm(crm)

Arguments

Details

To assure equalfinality the sill parameter for spatially correlated random residuals is fixed and standardized to 1.

Value

An object of a class "variogramModel" extending data.frame.

Author(s)

Kasia Sawicka, Gerard Heuvelink

Define Mulivariate Uncertainty Model

Description

Function that uses output of defineUM() to define joint probability distribution for uncertain cross-correlated variables.

Usage

defineMUM(UMlist, cormatrix, ...)

Arguments

Details

The cormatrix is a square matrix of correlations, dimensionally equal to the number of objects, symmetrical (transposed must be the same as original), diagonal must all be 1 all values must be <-1, +1>) and all eigenvalues must be > 0. The marginal Um objects must have provided id.

Value

Object of a class "JointNumericSpatial" or "JointScalar".

Author(s)

Kasia Sawicka, Gerard Heuvelink

Examples

set.seed(12345)

data(OC, OC_sd, TN, TN_sd)
OC_crm <- makeCRM(acf0 = 0.6, range = 5000, model = "Sph")
OC_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(OC, OC_sd), crm = OC_crm, id = "OC")
class(OC_UM)
TN_crm <- makeCRM(acf0 = 0.4, range = 5000, model = "Sph")
TN_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(TN, TN_sd), crm = TN_crm, id = "TN")
class(TN_UM)

soil_prop <- list(OC_UM,TN_UM)
mySpatialMUM <- defineMUM(soil_prop, matrix(c(1,0.7,0.7,1), nrow=2, ncol=2))
class(mySpatialMUM)

# scalar
scalarUM <- defineUM(uncertain = TRUE, distribution = "norm", 
                     distr_param = c(1, 2), id="Var1")                
scalarUM2 <- defineUM(uncertain = TRUE, distribution = "norm",
                      distr_param = c(3, 2), id="Var2")
scalarUM3 <- defineUM(uncertain = TRUE, distribution = "norm",
                      distr_param = c(10, 2.5), id="Var3")                
myMUM <- defineMUM(UMlist = list(scalarUM, scalarUM2, scalarUM3), 
               matrix(c(1,0.7,0.2,0.7,1,0.5,0.2,0.5,1), nrow = 3, ncol = 3))
class(myMUM)

Define an uncertainty model for a single variable

Description

Function that allows to define marginal uncertainty distributions for model inputs and subsequent Monte Carlo analysis.

Usage

defineUM(
  uncertain = TRUE,
  distribution = NULL,
  distr_param = NULL,
  crm = NULL,
  categories = NULL,
  cat_prob = NULL,
  id = NULL,
  ...
)

Arguments

Details

If the uncertain object is a spatial object, the distribution parameters or the probabilities for categories must be provided by means of maps, for example if a spatial variable has a normal distribution, a map of means and standard deviations must be provided. If crm is provided and spatial correlation between the residuals is assumed only the normal distribution for residuals is allowed.

If no spatial correlations between residuals is assumed, allowed distributions for marginal uncertainty models are listed in Table 1.

Table 1 Parametric probability models allowed in defineUM(). For more details look up ?distribution.

Value

Object of a class "MarginalXxx" that includes all necessary information for creating realizations of the uncertain variable. If provided arguments are: type of the distribution and corresponding parameters, and corresponding parameters are spatial objects - an object of class "MarginalNumericSpatial". If provided arguments are: type of the distribution and corresponding parameters, and corresponding parameters are non-spatial objects - an object of class "MarginalNumericSpatial". If provided arguments are: categories and probabilities, and probabilities are saved in a spatial object - an object of class "MarginalCategoricalSpatial". If provided arguments are: categories and probabilities, and probabilities are saved in a non-spatial object - an object of class "MarginalCategoricalDataFrame".

Author(s)

Kasia Sawicka, Gerard Heuvelink

Examples

# define uncertainty model for spatial numerical variable
data(dem30m, dem30m_sd)
dem_crm <- makeCRM(acf0 = 0.78, range = 321, model = "Exp")
demUM <- defineUM(uncertain = TRUE, distribution = "norm",
                   distr_param = c(dem30m, dem30m_sd), crm = dem_crm)
class(demUM)
                   
# define uncertainty model for spatial categorical variable
data(woon)
woonUM <- defineUM(TRUE, categories = c(1,2,3), cat_prob = woon[, c(4:6)])
class(woonUM)

# define uncertainty model for a variable desribed by a scalar
scalarUM <- defineUM(uncertain = TRUE, distribution = "gamma", distr_param = c(1,2))
class(scalarUM)

# define uncertainty model for two spatial cross-correlated variables
data(OC, OC_sd, TN, TN_sd)

OC_crm <- makeCRM(acf0 = 0.6, range = 1000, model = "Sph")
OC_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(OC, OC_sd), crm = OC_crm, id = "OC")
class(OC_UM)

TN_crm <- makeCRM(acf0 = 0.4, range = 1000, model = "Sph")
TN_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(TN, TN_sd), crm = TN_crm, id = "TN")
class(TN_UM)

Digital Elevation Model of Zlatibor region in Serbia.

Description

A dataset containing the mean an example Digital Elevation Model.

Usage

data(dem30m)

Format

a SpatialGridDataFrame with 15000 rows and 1 variable:

Elevation

Digital Elevation Model, in meters

Source

The Zlatibor dataset was kindly provided by Prof. Branislav Bajat from the University of Belgrade, Serbia.

Standard deviation of Digital Elevation Model of Zlatibor region in Serbia.

Description

A dataset containing the sd of an example Digital Elevation Model. It was calculated from dem30m using terrain funtion from raster package (opt = 'roughness').

Usage

data(dem30m_sd)

Format

a SpatialGridDataFrame with 15000 rows and 1 variable:

Elevation_sd

Standard deviation of Digital Elevation Model, in meters

Source

The Zlatibor dataset was kindly provided by Prof. Branislav Bajat from the University of Belgrade, Serbia.

Sampling from a given distribution

Description

Sampling from a given distribution

Usage

distribution_sampling(n, distribution, parameters)

Arguments

Value

Sample of random deviates.

Author(s)

Kasia Sawicka

Sampling from a given distribution

Description

Only used in samplemethod "randomSampling" for MarginalNumericSpatial.

Usage

distribution_sampling_raster(distribution, parameters_stack)

Arguments

Value

Sample of random deviates.

Author(s)

Kasia Sawicka

Wrapper function for calling executables in R

Description

Wrapper function for calling executables in R

Usage

executable(filename)

Arguments

Value

Executable output.

Author(s)

Dennis Walvoort

Sampling from a given distribution

Description

Sampling from a given distribution

Usage

find_strata(p, distribution, parameters, ...)

Arguments

Value

Strata of the distribution defined by given quantiles.

Author(s)

Kasia Sawicka, Stefan van Dam

Methods for generating Monte Carlo realizations from uncertain inputs.

Description

Methods for classes: "MarginalNumericSpatial", "MarginalScalar", "MarginalCategoricalSpatial", "JointNumericSpatial", "JointScalar". Function that runs Monte Carlo simulations depending on the type of uncertain object. Facilitates unconditional Gaussian simulation of errors for spatially auto-correlated residuals, as well as random and stratified random sampling if no spatial auto-correlation is included.

Usage

genSample(
  UMobject,
  n,
  samplemethod,
  p = 0,
  asList = TRUE,
  debug.level = 1,
  ...
)

Arguments

Details

Sampling methods:

"ugs" Unconditional Gaussian simulation of spatially auto-correlated and/or cross-correlated errors.

"randomSampling" Sampling multivariate distribution using eigenvalue decomposition (based on 'mvtnorm' package).

"stratifiedSampling" Number of samples (n) must be dividable by the number of quantiles to assure that each quantile is evenly represented.

"lhs" Not implemented yet. Sampling method for at least two uncertain inputs. The uncertain.object is then a list of two or more. It uses a stratified sampling method to generate inputs for the latin hypercube algorithm.

NOTE. Version 1.3-1 includes bug fixing related to derivation of cross-correlation matrix for multivariate uncertainty propagation analysis.

Value

A Monte Carlo sample of the variables of interest. If asList = TRUE returns list of all samples as lists.

Author(s)

Kasia Sawicka, Stefan van Dam, Gerard Heuvelink

Examples

set.seed(12345)

### ------------------- "MarginalNumericSpatial" -------------------
# load data
data(dem30m, dem30m_sd)

# "ugs" method example
dem_crm <- makeCRM(acf0 = 0.78, range = 321, model = "Exp")
demUM <- defineUM(uncertain = TRUE, distribution = "norm", 
                   distr_param = c(dem30m, dem30m_sd), crm = dem_crm)
                   
# toy example
dem_sample <- genSample(UMobject = demUM, n = 2, samplemethod = "ugs", nmax = 4, asList = FALSE)
str(dem_sample)
# any meaningful Monte Carlo analysis should have normally much larger number of runs               
## Not run: 
dem_sample <- genSample(UMobject = demUM, n = 100, samplemethod = "ugs", nmax = 20, asList = FALSE)
str(dem_sample)

## End(Not run)

# "startifiedSampling" method example
demUM <- defineUM(uncertain = TRUE, distribution = "norm", distr_param = c(dem30m, dem30m_sd))
# toy example
dem_sample <- genSample(UMobject = demUM, n = 5, samplemethod = "stratifiedSampling", p = 0:5/5)
# any meaningful Monte Carlo analysis should have normally much larger number of runs
## Not run: 
dem_sample <- genSample(UMobject = demUM, n = 100, samplemethod = "stratifiedSampling", p = 0:5/5)
str(dem_sample)

## End(Not run)

# Examples with rasters
# (raster with auto-correlation)
data(OC, OC_sd)
OC_crm <- makeCRM(acf0 = 0.6, range = 1000, model = "Sph")
OC_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(OC, OC_sd), crm = OC_crm, id = "OC")
class(OC_UM)
# toy example
some_sample <- genSample(OC_UM, n = 2, "ugs", nmax = 4)
some_sample
# any meaningful Monte Carlo analysis should have normally much larger number of runs
## Not run: 
some_sample <- genSample(OC_UM, n = 50, "ugs", nmax = 24)
some_sample

## End(Not run)


### ----------------------- "MarginalScalar" -----------------------
# example with normal distribution
scalarUM <- defineUM(uncertain = TRUE, distribution = "norm", distr_param = c(10, 1))
scalar_sample <- genSample(scalarUM, n = 10, samplemethod = "randomSampling")


### ----------------- "MarginalCategoricalSpatial" -----------------
# load data
data(woon)
woonUM <- defineUM(TRUE, categories = c(1,2,3), cat_prob = woon[, c(4:6)])
woon_sample <- genSample(woonUM, 10, asList = FALSE)
class(woon_sample)
str(woon_sample@data)


### -------------------- "JointNumericSpatial" ----------------------
# load data
data(OC, OC_sd, TN, TN_sd)

# define marginal UMs
OC_crm <- makeCRM(acf0 = 0.6, range = 1000, model = "Sph")
OC_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(OC, OC_sd), crm = OC_crm, id = "OC")
TN_crm <- makeCRM(acf0 = 0.4, range = 1000, model = "Sph")
TN_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(TN, TN_sd), crm = TN_crm, id = "TN")

# define joint UM
soil_prop <- list(OC_UM, TN_UM)
mySpatialMUM <- defineMUM(soil_prop, matrix(c(1,0.7,0.7,1), nrow=2, ncol=2))
class(mySpatialMUM)

# sample - "ugs" method
# toy example
my_cross_sample <- genSample(mySpatialMUM, n = 2, "ugs", nmax = 4, asList = TRUE)
class(my_cross_sample)
# any meaningful Monte Carlo analysis should have normally much larger number of runs
## Not run: 
my_cross_sample <- genSample(mySpatialMUM, n = 100, "ugs", nmax = 24, asList = TRUE)
class(my_cross_sample)

## End(Not run)


### ------------------------- "JointScalar" ------------------------
scalarUM <- defineUM(uncertain = TRUE, distribution = "norm",
                     distr_param = c(1, 2), id="Var1")                
scalarUM2 <- defineUM(uncertain = TRUE, distribution = "norm",
                      distr_param = c(3, 2), id="Var2")
scalarUM3 <- defineUM(uncertain = TRUE, distribution = "norm",
                      distr_param = c(10, 2.5), id="Var3")                
myMUM <- defineMUM(UMlist = list(scalarUM, scalarUM2, scalarUM3), 
               matrix(c(1,0.7,0.2,0.7,1,0.5,0.2,0.5,1), nrow = 3, ncol = 3))
my_sample <- genSample(myMUM, n = 5, samplemethod = "randomSampling", asList = FALSE)
my_sample  

Generating Monte Carlo sample from a list of uncertain objects that are cross-correlated.

Description

Uncertain objects are described by joint PDF or a list from independent objects. Sampling can be done via three different sampling methods:

Usage

## S3 method for class 'JointNumericSpatial'
genSample(
  UMobject,
  n,
  samplemethod,
  p = 0,
  asList = TRUE,
  debug.level = 1,
  ...
)

Arguments

Details

"ugs" Unconditional gaussian simulation of spatially cross-correlated errors.

"randomSampling" Sampling multivariate distribution using eigenvalue decomposition (based on 'mvtnorm' package).

"lhs" Not implemented yet. Sampling method for at least two uncertain inputs. The uncertain.object is then a list of two or more. It uses startified sampling method to generate the inputs for the latin hypercude algorithm, hence number of samples (n) must be dividable by the number of quantiles to assure each quantile is evenly represented.

NOTE. Version 1.3-1 includes bug fixing related to derivation of cross-correlation matrix for multivariate uncertainty propagation analysis.

Value

A Monte Carlo sample of the variables of interest. If asList = TRUE returns list of all samples as lists.

Author(s)

Kasia Sawicka, Stefan van Dam, Gerard Heuvelink

Examples

set.seed(12345)
# "ugs" method example
# load data
data(OC, OC_sd, TN, TN_sd)

# Test for SpatialGridDataFrames
OC <- as(OC, 'SpatialGridDataFrame')
TN <- as(TN, 'SpatialGridDataFrame')
OC_sd <- as(OC_sd, 'SpatialGridDataFrame')
TN_sd <- as(TN_sd, 'SpatialGridDataFrame')

# define marginal UMs
OC_crm <- makeCRM(acf0 = 0.6, range = 5000, model = "Sph")
OC_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(OC, OC_sd), crm = OC_crm, id = "OC")
TN_crm <- makeCRM(acf0 = 0.4, range = 5000, model = "Sph")
TN_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(TN, TN_sd), crm = TN_crm, id = "TN")

# define joint UM
soil_prop <- list(OC_UM, TN_UM)
mySpatialMUM <- defineMUM(soil_prop, matrix(c(1,0.7,0.7,1), nrow=2, ncol=2))

# sample - "ugs" method
# toy example
my_cross_sample <- genSample(mySpatialMUM, n = 3, "ugs", nmax = 24, asList = TRUE)
class(my_cross_sample)
## Not run: 
my_cross_sample <- genSample(mySpatialMUM, n = 50, "ugs", nmax = 24, asList = TRUE)
class(my_cross_sample)

## End(Not run)

Generating sample from cross-correlated variables described by a scalar.

Description

Generating sample from cross-correlated variables described by a scalar.

Usage

## S3 method for class 'JointScalar'
genSample(UMobject, n, samplemethod, p = 0, asList = TRUE, ...)

Arguments

Value

Monte Carlo sample of cross-correlated scalar variables.

Author(s)

Kasia Sawicka, Gerard Heuvelink

Examples

set.seed(12345)
scalarUM <- defineUM(uncertain = TRUE, distribution = "norm",
                     distr_param = c(1, 2), id="Var1")                
scalarUM2 <- defineUM(uncertain = TRUE, distribution = "norm",
                      distr_param = c(3, 2), id="Var2")
scalarUM3 <- defineUM(uncertain = TRUE, distribution = "norm",
                      distr_param = c(10, 2.5), id="Var3")                
myMUM <- defineMUM(UMlist = list(scalarUM, scalarUM2, scalarUM3), 
               matrix(c(1,0.7,0.2,0.7,1,0.5,0.2,0.5,1), nrow = 3, ncol = 3))
my_sample <- genSample(myMUM, n = 10, samplemethod = "randomSampling", asList = FALSE)
my_sample  

Generating Monte Carlo sample from an uncertain object of a class 'MarginalCategoricalSpatial'

Description

Generating Monte Carlo sample from an uncertain object of a class 'MarginalCategoricalSpatial'

Usage

## S3 method for class 'MarginalCategoricalSpatial'
genSample(UMobject, n, samplemethod, p = 0, asList = TRUE, ...)

Arguments

Value

A Monte Carlo sample of a categorical spatial variable.

Author(s)

Kasia Sawicka

Examples

set.seed(12345)
# load data
data(woon)
woonUM <- defineUM(TRUE, categories = c(1,2,3), cat_prob = woon[, c(4:6)])
woon_sample <- genSample(woonUM, 10, asList = FALSE)
class(woon_sample)
str(woon_sample@data)
woon_sample <- genSample(woonUM, 10)
class(woon_sample)

# analyse probability of having snow
# load data
data(dem30m, dem30m_sd)

# generate dummy probabilities for categories "snow" and "no snow"
dem30m$snow_prob <- NA
dem30m$snow_prob[dem30m$Elevation > 1000] <- 0.75
dem30m$snow_prob[dem30m$Elevation <= 1000] <- 0.25
dem30m$no_snow_prob <- 1 - dem30m$snow_prob
summary(dem30m@data)
snowUM <- defineUM(uncertain = TRUE, categories = c("snow", "no snow"), cat_prob = dem30m[2:3])
class(snowUM)
snow_sample <- genSample(snowUM, 10, asList = FALSE)
head(snow_sample@data)

Generating Monte Carlo sample from an uncertain object of a class 'MarginalNumericSpatial'

Description

Function that runs Monte Carlo simulations depending on the type of uncertain object. Facilitates unconditional gausian simulation of errors for spatially auto-correlated residulas, and random sampling, stratified sampling if no spatial auto-correlation is included.

Usage

## S3 method for class 'MarginalNumericSpatial'
genSample(
  UMobject,
  n,
  samplemethod,
  p = 0,
  asList = TRUE,
  debug.level = 1,
  ...
)

Arguments

Details

"ugs" Unconditional gaussian simulation of spatially auto-correlated errors.

"stratifiedSampling" Number of samples (n) must be dividable by the number of quantiles to assure each quantile is evenly represented.

"lhs" Sampling method for at least two uncertain inputs. The uncertain.object is then a list of two or more. It uses startified sampling method to generate the inputs for the latin hypercude algorithm, hence the p is restricted as above.

Value

A Monte Carlo sample of uncertain input of a class of distribution parameters.

Author(s)

Kasia Sawicka, Stefan van Dam, Gerard Heuvelink

Examples

set.seed(12345)
# load data
data(dem30m, dem30m_sd)

# "ugs" method example
dem_crm <- makeCRM(acf0 = 0.78, range = 321, model = "Exp")
demUM <- defineUM(uncertain = TRUE, distribution = "norm", 
                   distr_param = c(dem30m, dem30m_sd), crm = dem_crm)
                   
# toy example
dem_sample <- genSample(UMobject = demUM, n = 2, samplemethod = "ugs", nmax = 6, asList = FALSE)
str(dem_sample)               
## Not run: 
dem_sample <- genSample(UMobject = demUM, n = 50, samplemethod = "ugs", nmax = 20, asList = FALSE)
str(dem_sample)

## End(Not run)


# "startifiedSampling" method example
demUM <- defineUM(uncertain = TRUE, distribution = "norm", distr_param = c(dem30m, dem30m_sd))
# toy example
dem_sample <- genSample(UMobject = demUM, n = 5, samplemethod = "stratifiedSampling", p = 0:5/5)
## Not run: 
dem_sample <- genSample(UMobject = demUM, n = 50, samplemethod = "stratifiedSampling", p = 0:5/5)
str(dem_sample)

## End(Not run)

# Examples with rasters
# (raster with auto-correlation)
data(OC, OC_sd)
OC_crm <- makeCRM(acf0 = 0.6, range = 1000, model = "Sph")
OC_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(OC, OC_sd), crm = OC_crm, id = "OC")
class(OC_UM)
# toy example
some_sample <- genSample(OC_UM, n = 3, "ugs", nmax=6)
some_sample
## Not run: 
some_sample <- genSample(OC_UM, n = 50, "ugs", nmax=20)
some_sample

## End(Not run)

Generating Monte Carlo sample from an uncertain object of a class 'MarginalScalar'

Description

Function that runs Monte Carlo simulations for MarginalScalar class objects.

Usage

## S3 method for class 'MarginalScalar'
genSample(UMobject, n, samplemethod, p = 0, asList = TRUE, ...)

Arguments

Details

"stratifiedSampling" Number of samples (n) must be dividable by the number of quantiles to assure each quantile is evenly represented.

Value

A Monte Carlo sample of uncertain input of a class of distribution parameters.

Author(s)

Kasia Sawicka

Examples

set.seed(12345)
# Example 1
scalarUM <- defineUM(uncertain = TRUE, distribution = "norm", distr_param = c(10, 1))
scalar_sample <- genSample(scalarUM, n = 10, samplemethod = "randomSampling")

# Example 2
scalarUM <- defineUM(uncertain = TRUE, distribution = "beta", distr_param = c(10, 1, 2))
scalar_sample <- genSample(scalarUM, n = 10, samplemethod = "stratifiedSampling", p = 0:5/5)

Function to find the level of list nesting

Description

Function to find the level of list nesting

Usage

list_depth(List)

Arguments

Value

an integer; level of list nesting

Author(s)

Kasia Sawicka

Examples

a <- list(1,2)
list_depth(a)

a <- list(list(1, 2), 3)
list_depth(a)

Defining a spatial correlogram model

Description

Function that generates a spatial correlogram model, an object of class "SpatialCorrelogramModel".

Usage

makeCRM(
  acf0 = 1,
  range = NA,
  model,
  anis,
  kappa = 0.5,
  add.to,
  covtable,
  Err = 0
)

Arguments

Details

For the spatial variables allowed autocorrelation functions are listed in Table 4.1 of the gstat manual. Spatial correlation assumes stationarity, i.e. correlation depends only on the separation distance between points in space. Anisotropy is allowed. No nested models are allowed in the current version.

Value

An object of a class "SpatialCorrelogramModel". This is a list collating provided arguments.

Author(s)

Kasia Sawicka, Gerard Heuvelink

Examples

mycormodel <- makeCRM(acf0 = 0.8, range = 300, model = "Exp")
str(mycormodel)

mean() function for MC sample saved in a SpatialGridDataFrame

Description

Calculates mean from MC realizations for each location in a map.

Usage

mean_MC_sgdf(realizations, ...)

Arguments

Value

SpatialGridDataFrame; a mean of a MC sample.

Author(s)

Kasia Sawicka

Examples

set.seed(12345)
# load data
data(dem30m, dem30m_sd)
dem_crm <- makeCRM(acf0 = 0.78, range = 321, model = "Exp")
demUM <- defineUM(uncertain = TRUE, distribution = "norm", 
                   distr_param = c(dem30m, dem30m_sd), crm = dem_crm)
dem_sample <- genSample(UMobject = demUM, n = 50, samplemethod = "ugs", 
                        nmax = 20, asList = FALSE)
dem_mean <- mean_MC_sgdf(dem_sample)

Plots correlogram model

Description

Plots correlogram model

Usage

## S3 method for class 'SpatialCorrelogramModel'
plot(
  x,
  distance = 1,
  ylim = c(0, 1),
  xlab = "Distance",
  ylab = "Correlation",
  ...
)

Arguments

Value

plot of correlogram model

Author(s)

Kasia Sawicka, Gerard Heuvelink

Examples

mycormodel <- makeCRM(acf0 = 0.8, range = 300, model = "Exp")
plot(mycormodel, distance = 1)

Print method for class "template."

Description

Print method for class "template."

Usage

## S3 method for class 'template'
print(x, ...)

Arguments

Value

Template file content.

Author(s)

Dennis Walvoort

Propagation function

Description

A function that runs a model repeatedly with Monte Carlo samples of uncertain inputs.

Usage

propagate(realizations, model, n, ...)

Arguments

Value

Model output Monte Carlo realizations.

Author(s)

Kasia Sawicka

Examples

set.seed(12345)
## continuous spatial data example with a single variable
# load data
data(dem30m, dem30m_sd)

# Slope model 
Slope <- function(DEM, ...) {
  require(raster)
  require(purrr)
  demraster <- 
    DEM %>%
    raster()
  demraster %>%
    terrain(opt = 'slope', ...) %>%
    as("SpatialGridDataFrame")
}
 
# uncertainty propagation
dem_crm <- makeCRM(acf0 = 0.78, range = 321, model = "Exp")
demUM <- defineUM(uncertain = TRUE, distribution = "norm", 
                   distr_param = c(dem30m, dem30m_sd), crm = dem_crm)
# toy example
dem_sample <- genSample(UMobject = demUM, n = 3, samplemethod = "ugs", nmax = 20)
slope_sample <- propagate(dem_sample, model = Slope, n = 3)
## Not run: 
dem_sample <- genSample(UMobject = demUM, n = 50, samplemethod = "ugs", nmax = 20)
slope_sample <- propagate(dem_sample, model = Slope, n = 50)

## End(Not run)

## categorical spatial data example
# load data
data(woon)

# tax model
tax <- function(building_Function) { 
  building_Function$tax2pay <- NA
  building_Function$tax2pay[building_Function$Function == 1] <- 1000
  building_Function$tax2pay[building_Function$Function == 2] <- 10000
  building_Function$tax2pay[building_Function$Function == 3] <- 10
  total_tax <- sum(building_Function$tax2pay)
  total_tax
}

# uncertainty propagation
woonUM <- defineUM(TRUE, categories = c(1,2,3), cat_prob = woon[, c(4:6)])
woon_sample <- genSample(woonUM, 10)
class(woon_sample)
tax # the model takes SpatialGridDataFrame with a column called "Function"
for (i in 1:10) names(woon_sample[[i]]) <- "Function"
tax_uncert <- propagate(realizations = woon_sample, n = 10, model = tax)
tax_uncert <- unlist(tax_uncert)
summary(tax_uncert)

## cross-correlated example
# load data
data(OC, OC_sd, TN, TN_sd)

# C/N model
C_N_model_raster <- function(OC, TN) {
  OC/TN
}

# define marginal UMs
OC_crm <- makeCRM(acf0 = 0.6, range = 1000, model = "Sph")
OC_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(OC, OC_sd), crm = OC_crm, id = "OC")
TN_crm <- makeCRM(acf0 = 0.4, range = 1000, model = "Sph")
TN_UM <- defineUM(TRUE, distribution = "norm", distr_param = c(TN, TN_sd), crm = TN_crm, id = "TN")

# define joint UM
mySpatialMUM <- defineMUM(list(OC_UM, TN_UM), matrix(c(1,0.7,0.7,1), nrow=2, ncol=2))

# sample - "ugs" method
# toy example
my_cross_sample <- genSample(mySpatialMUM, n = 3, "ugs", nmax = 24)
class(my_cross_sample)
# run propagation
CN_sample <- propagate(realizations = my_cross_sample, model = C_N_model_raster, n = 3)
CN_sample
## Not run: 
my_cross_sample <- genSample(mySpatialMUM, 50, "ugs", nmax = 24)
class(my_cross_sample)
# run propagation
CN_sample <- propagate(realizations = my_cross_sample, model = C_N_model_raster, n = 50)
CN_sample

## End(Not run)

quantile() function for MC sample saved in a SpatialGridDataFrame

Description

Calculates mean from MC realizations for each location in a map.

Usage

quantile_MC_sgdf(realizations, ...)

Arguments

Value

SpatialGridDataFrame; quantiles of a MC sample

Author(s)

Kasia Sawicka

Examples

set.seed(12345)
data(dem30m, dem30m_sd)
dem_crm <- makeCRM(acf0 = 0.78, range = 321, model = "Exp")
demUM <- defineUM(uncertain = TRUE, distribution = "norm", 
                   distr_param = c(dem30m, dem30m_sd), crm = dem_crm)
## Not run: 
dem_sample <- genSample(UMobject = demUM, n = 50, samplemethod = "ugs",
                        nmax = 20, asList = FALSE)
dem_quantile <- quantile_MC_sgdf(dem_sample, probs = c(0.1, 0.9))

## End(Not run)

Rendering template

Description

Rendering is the process of replacing the tags in moustaches by text. For this, we provide a set of render-methods. See the 'whisker' package (or https://mustache.github.io) for more information.

Usage

render(x, ...)

Arguments

Value

Rendered character template or a file on disk.

Author(s)

Dennis Walvoort

Examples

require(magrittr)
require(whisker)
# render character string
my_template <- "Hello {{name}}. How are you doing?"
my_template %>% 
  render(name = "Winnie the Pooh")

# render table      
my_template <- c(
     "| x | y |",
     "|---|---|",
     "{{#MY_TABLE}}",
     "| {{X}} | {{Y}} |",
     "{{/MY_TABLE}}")
my_table <- data.frame(X = 1:5, Y = letters[1:5])  
my_table
my_template %>% 
render(MY_TABLE = unname(rowSplit(my_table))) %>%
cat

Render method for "character" class.

Description

Rendering is the process of replacing the tags in moustaches by text.

Usage

## S3 method for class 'character'
render(x, ...)

Arguments

Value

Rendered character template.

Author(s)

Dennis Walvoort

Examples

require(magrittr)
require(whisker)
# render character string
my_template <- "Hello {{name}}. How are you doing?"
my_template %>% 
  render(name = "Winnie the Pooh")

# render table      
my_template <- c(
     "| x | y |",
     "|---|---|",
     "{{#MY_TABLE}}",
     "| {{X}} | {{Y}} |",
     "{{/MY_TABLE}}")
my_table <- data.frame(X = 1:5, Y = letters[1:5])  
my_table
my_template %>% 
render(MY_TABLE = unname(rowSplit(my_table))) %>%
cat
  

Render method for "template" class.

Description

Rendering is the process of replacing the tags in moustaches by text.

Usage

## S3 method for class 'template'
render(x, ...)

Arguments

Value

Rendered template file.

Author(s)

Dennis Walvoort

sd() function for MC sample saved in a SpatialGridDataFrame

Description

Calculates sd from MC realizations for each location in a map.

Usage

sd_MC_sgdf(realizations, ...)

Arguments

Value

SpatialGridDataFrame; a sd of a MC sample.

Author(s)

Kasia Sawicka

Examples

set.seed(12345)
data(dem30m, dem30m_sd)
dem_crm <- makeCRM(acf0 = 0.78, range = 321, model = "Exp")
demUM <- defineUM(uncertain = TRUE, distribution = "norm", 
                   distr_param = c(dem30m, dem30m_sd), crm = dem_crm)
## Not run: 
dem_sample <- genSample(UMobject = demUM, n = 50, samplemethod = "ugs",
                        nmax = 20, asList = FALSE)
dem_sample_sd <- sd_MC_sgdf(dem_sample)

## End(Not run)

spup - Package for spatial uncertainty propagation

Description

Facilitates uncertainty propagation analysis using Monte Carlo methods. In particular, provides functions that allow to do uncertainty analysis with spatial variables/models.

Author(s)

Kasia Sawicka

Stratified sampling for spatial variables

Description

Stratified sampling for spatial variables

Usage

stratsamp(n, distribution, parameters, p)

Arguments

Value

Sample of spatial variable. Matrix with n rows and length(p)-1 columns.

Author(s)

Stefan van Dam, Kasia Sawicka

Constructor for class "template".

Description

Class that stores all templates with model inputs. The aim of this class is to: 1. organise model input files; 2. perform some checks.

Usage

template(filenames)

Arguments

Details

A template is a model input file with: 1. the additional extension '.template'. 2. input that needs to be modified is replaced by mustache-style tags.

Value

An object of a class "template".

Author(s)

Dennis Walvoort

var() function for MC sample saved in a SpatialGridDataFrame

Description

Calculates var from MC realizations for each location in a map.

Usage

var_MC_sgdf(realizations, ...)

Arguments

Value

SpatialGridDataFrame; a variance of a MC sample.

Author(s)

Kasia Sawicka

Examples

set.seed(12345)
data(dem30m, dem30m_sd)
dem_crm <- makeCRM(acf0 = 0.78, range = 321, model = "Exp")
demUM <- defineUM(uncertain = TRUE, distribution = "norm", 
                   distr_param = c(dem30m, dem30m_sd), crm = dem_crm)
## Not run: 
dem_sample <- genSample(UMobject = demUM, n = 50, samplemethod = "ugs",
                        nmax = 20, asList = FALSE)
dem_var <- var_MC_sgdf(dem_sample)

## End(Not run)

Calculate variance covariance matrix

Description

Calculate variance covariance matrix

Usage

varcov(sd_vector, cormat)

Arguments

Value

Variance-covariance matrix.

Author(s)

Kasia Sawicka

Examples

vc <- varcov(c(1,2,3), matrix(c(1,0.7,0.2,0.7,1,0.5,0.2,0.5,1), nrow = 3, ncol = 3))
vc

Convert vgm to crm

Description

Convert vgm to crm

Usage

vgm2crm(vgm, psill, nugget, range, model, kappa = 0.5, Err = 0)

Arguments

Value

Spatial correlogram model - standardised parameters of spatial variogram model

Author(s)

Kasia Sawicka

Neighbourhood in Rotterdam.

Description

The 'woon' object is a SpatialPolygonDataFrame where each building is represented by one polygon.

Usage

data(woon)

Format

a SpatialPolygonDataFrame with 723 polygons and 7 variables:

vbos

number of addresses present in the building

woonareash

residential area, in percent

Function

assigned category depending on vbos and woonareash - for residential is 1, for offcie is 2, for other is 3

residential

probability that the building is residential

office

probability that the building is an office

other

probability that the building has other function

check

check if probabilities sum to 1

Source

Kadaster, NL.