| Type: | Package |
| Title: | Extension to 'spatstat' for Large Datasets on a Linear Network |
| Version: | 3.1-2 |
| Date: | 2024-09-05 |
| Depends: | R (≥ 3.5.0), spatstat.data (≥ 3.1-0), spatstat.sparse (≥ 3.1-0), spatstat.univar (≥ 3.0-0), spatstat.geom (≥ 3.3-0), spatstat.random (≥ 3.3-0), spatstat.explore (≥ 3.3-0), spatstat.model (≥ 3.3-0), spatstat.linnet (≥ 3.2-0), spatstat (≥ 3.1-1) |
| Imports: | spatstat.utils (≥ 3.0-5), Matrix |
| Maintainer: | Adrian Baddeley <Adrian.Baddeley@curtin.edu.au> |
| Description: | Extension to the 'spatstat' family of packages, for analysing large datasets of spatial points on a network. The geometrically- corrected K function is computed using a memory-efficient tree-based algorithm described by Rakshit, Baddeley and Nair (2019). |
| License: | GPL-2 | GPL-3 [expanded from: GPL (≥ 2)] |
| NeedsCompilation: | yes |
| ByteCompile: | true |
| Packaged: | 2024-09-05 07:36:23 UTC; adrian |
| Author: | Suman Rakshit 
[aut, cph],
Adrian Baddeley
[cre, cph] |
| Repository: | CRAN |
| Date/Publication: | 2024-09-05 14:50:02 UTC |
Extension to 'spatstat' for Large Datasets on a Linear Network
Description
Extension to the 'spatstat' family of packages, for analysing large datasets of spatial points on a network. The geometrically- corrected K function is computed using a memory-efficient tree-based algorithm described by Rakshit, Baddeley and Nair (2019).
Details
This is an extension to the spatstat package for the analysis of large data sets on linear networks.
Its main functionality is a memory-efficient algorithm
for computing the estimate of the K function
on a linear network, described in Rakshit et al (2019).
The main functions are Knet and Knetinhom.
These are counterparts of the functions
linearK
and
linearKinhom
in the spatstat.linnet package.
The spatstat.linnet functions
linearK
and
linearKinhom
are usable (and slightly faster)
for small datasets, but require substantial amounts of memory.
For larger datasets,
the functions Knet and Knetinhom
are much more efficient.
The DESCRIPTION file:
| Package: | spatstat.Knet |
| Type: | Package |
| Title: | Extension to 'spatstat' for Large Datasets on a Linear Network |
| Version: | 3.1-2 |
| Date: | 2024-09-05 |
| Depends: | R (>= 3.5.0), spatstat.data (>= 3.1-0), spatstat.sparse (>= 3.1-0), spatstat.univar (>= 3.0-0), spatstat.geom (>= 3.3-0), spatstat.random (>= 3.3-0), spatstat.explore (>= 3.3-0), spatstat.model (>= 3.3-0), spatstat.linnet (>= 3.2-0), spatstat (>= 3.1-1) |
| Imports: | spatstat.utils (>= 3.0-5), Matrix |
| Authors@R: | c(person(given="Suman", family="Rakshit", role = c("aut", "cph"), email = "suman.rakshit@curtin.edu.au", comment=c(ORCID="0000-0003-0052-128X")), person(given="Adrian", family="Baddeley", role = c("cre", "cph"), email = "Adrian.Baddeley@curtin.edu.au", comment = c(ORCID="0000-0001-9499-8382"))) |
| Maintainer: | Adrian Baddeley <Adrian.Baddeley@curtin.edu.au> |
| Description: | Extension to the 'spatstat' family of packages, for analysing large datasets of spatial points on a network. The geometrically- corrected K function is computed using a memory-efficient tree-based algorithm described by Rakshit, Baddeley and Nair (2019). |
| License: | GPL (>= 2) |
| NeedsCompilation: | yes |
| ByteCompile: | true |
| Author: | Suman Rakshit [aut, cph] (<https://orcid.org/0000-0003-0052-128X>), Adrian Baddeley [cre, cph] (<https://orcid.org/0000-0001-9499-8382>) |
Index of help topics:
Knet Geometrically-Corrected K Function on Network
Knetinhom Geometrically-Corrected Inhomogeneous K
Function on Network
spatstat.Knet-package Extension to 'spatstat' for Large Datasets on a
Linear Network
wacrashes Road Accidents in Western Australia
Author(s)
Suman Rakshit [aut, cph] (<https://orcid.org/0000-0003-0052-128X>), Adrian Baddeley [cre, cph] (<https://orcid.org/0000-0001-9499-8382>)
Maintainer: Adrian Baddeley <Adrian.Baddeley@curtin.edu.au>
References
Rakshit, S., Baddeley, A. and Nair, G. (2019)
Efficient code for second order analysis of events on a linear network.
Journal of Statistical Software 90 (1) 1–37.
DOI: 10.18637/jss.v090.i01
Geometrically-Corrected K Function on Network
Description
Compute the geometrically-corrected K function
for a point pattern on a linear network.
Usage
Knet(X, r = NULL, freq, ..., verbose=FALSE)
Arguments
X |
Point pattern on a linear network (object of class |
r |
Optional. Numeric vector of values of the function argument |
freq |
Vector of frequencies corresponding to the point events on the network. The length of this vector should be equal to the number of points on the network. The default frequency is one for every point on the network. |
... |
Ignored. |
verbose |
A logical for printing iteration number corresponding to each point event on the network. |
Details
This command computes the geometrically-corrected K function,
proposed by Ang et al (2012), from point pattern data on a linear
network.
The algorithm used in this computation is discussed in Rakshit et al (2019).
The spatstat function linearK
is usable (and slightly faster) for the same purpose
for small datasets, but requires substantial amounts of memory.
For larger datasets, the function Knet is much more efficient.
Value
Function value table (object of class "fv").
Author(s)
Suman Rakshit (modified by Adrian Baddeley)
References
Ang, Q.W., Baddeley, A. and Nair, G. (2012) Geometrically corrected second-order analysis of events on a linear network, with applications to ecology and criminology. Scandinavian Journal of Statistics 39, 591–617.
Rakshit, S., Baddeley, A. and Nair, G. (2019)
Efficient code for second order analysis of events on a linear network.
Journal of Statistical Software 90 (1) 1–37.
DOI: 10.18637/jss.v090.i01
Examples
UC <- unmark(chicago)
r <- seq(0, 1000, length = 41)
K <- Knet(UC, r = r)
Geometrically-Corrected Inhomogeneous K Function on Network
Description
Compute the geometrically-corrected inhomogeneous K function
for a point pattern on a linear network.
Usage
Knetinhom(X, lambda, r = NULL, freq, ..., verbose=FALSE)
Arguments
X |
Point pattern on a linear network (object of class |
lambda |
Fitted intensity of the point process. Either a numeric vector
giving values of the fitted intensity at each data point of
|
r |
Optional. Numeric vector of values of the function argument
|
freq |
Vector of frequencies corresponding to the point events on the network. The length of this vector should be equal to the number of points on the network. The default frequency is one for every point on the network. |
... |
Ignored. |
verbose |
Logical value indicating whether to print progress reports during the computation. |
Details
This command computes the inhomogeneous version of the
geometrically-corrected K function, proposed by Ang et al
(2012), from point pattern data on a linear network.
The algorithm used in this computation is described in Rakshit et al (2019).
The spatstat function
linearKinhom
is usable (and slightly faster) for this purpose
for small datasets, but requires substantial amounts of memory.
For larger datasets,
the function Knetinhom is much more efficient.
Value
Function value table (object of class "fv").
Author(s)
Suman Rakshit (modified by Adrian Baddeley)
References
Ang, Q.W., Baddeley, A. and Nair, G. (2012) Geometrically corrected second-order analysis of events on a linear network, with applications to ecology and criminology. Scandinavian Journal of Statistics 39, 591–617.
Rakshit, S., Baddeley, A. and Nair, G. (2019)
Efficient code for second order analysis of events on a linear network.
Journal of Statistical Software 90 (1) 1–37.
DOI: 10.18637/jss.v090.i01
Examples
UC <- unmark(chicago)
fit <- lppm(UC ~ x+y)
r <- seq(0, 1000, length = 41)
K <- Knetinhom(UC, lambda=fit, r = r)
Road Accidents in Western Australia
Description
This dataset gives the spatial locations of all road accidents recorded in the state of Western Australia for the year 2011, on the state road network.
These data were published and analysed in Rakshit et al (2019).
Usage
data(wacrashes)Format
A object of class "lpp" representing the spatial point pattern
of accident locations on the network of roads in Western Australia.
Details
The road network has 88,512 intersections and 115,169 road segments. The spatial coordinates are expressed in metres, and the total network length is 97,165,540 metres (97,165 km). The number of accident locations on the network is 14,562.
Source
Main Roads, Western Australia. Made available as part of the Western Australian Whole of Government Open Data Policy.
References
Rakshit, S., Baddeley, A. and Nair, G. (2019)
Efficient code for second order analysis of events on a linear network.
Journal of Statistical Software 90 (1) 1–37.
DOI: 10.18637/jss.v090.i01
Examples
data(wacrashes)
wacrashes
summary(wacrashes)
plot(wacrashes, cols="red", cex=0.5)