Package 'motif' reference manual

Title:	Local Pattern Analysis
Description:	Describes spatial patterns of categorical raster data for any defined regular and irregular areas. Patterns are described quantitatively using built-in signatures based on co-occurrence matrices but also allows for any user-defined functions. It enables spatial analysis such as search, change detection, and clustering to be performed on spatial patterns (Nowosad (2021) <doi:10.1007/s10980-020-01135-0>).
Authors:	Jakub Nowosad [aut, cre]
Maintainer:	Jakub Nowosad <[email protected]>
License:	MIT + file LICENSE
Version:	0.6.4
Built:	2025-02-19 05:15:49 UTC
Source:	https://github.com/nowosad/motif

Determine unique classes (internal function)

Description

Determine unique classes (internal function)

Usage

determine_classes(x, window)
determine_classes(x, window)

Arguments

`x`	a `stars` or `stars_proxy` object
`window`	a windows argument from `lsp_signature()`, `lsp_search()`, or `lsp_compare()`

Value

A list with vector of numbers (unique classes)

Adds clusters' ids to a lsp object

Description

Adds clusters' ids to a lsp object. The output can be of stars, sf, or terra class. See examples.

Usage

lsp_add_clusters(x, clust, output = "sf", window = NULL)
lsp_add_clusters(x, clust, output = "sf", window = NULL)

Arguments

`x`	Object of class `lsp` - usually the output of the `lsp_signature()` function
`clust`	Vector containing an id value for each row in `x`
`output`	The class of the output. Either `stars`, `sf`, or `terra`
`window`	Specifies areas for analysis. It can be either: `NULL` or an `sf` object. If `window=NULL` calculations are performed based on the metadata from `x`. If an `sf` object is provided, each feature (row) defines the extent of a local pattern. The `sf` object should have one attribute (otherwise, the first attribute is used as an id).

Value

Object of class stars, sf, or terra (depending on the output argument) with an additional column "clust" representing clusters' id values.

Examples


library(stars)
library(sf)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
landform_cove = lsp_signature(landform,
                               type = "cove",
                               window = 200,
                               normalization = "pdf")

landform_dist = lsp_to_dist(landform_cove,
                            dist_fun = "jensen-shannon")

landform_hclust = hclust(landform_dist, method = "ward.D2")
#plot(landform_hclust)

clusters = cutree(landform_hclust, k = 4)

landform_grid_sf = lsp_add_clusters(landform_cove, clusters)
#plot(landform_grid_sf["clust"])

#landform_grid_sfq = lsp_add_quality(landform_grid_sf,
#                                        landform_dist)
#plot(landform_grid_sfq["quality"])

## larger data example
# library(stars)
# library(sf)
# landform = read_stars(system.file("raster/landform.tif", package = "motif"),
#                       proxy = FALSE)
# landform_cove = lsp_signature(landform,
#                                type = "cove",
#                                window = 200,
#                                normalization = "pdf")
#
# landform_dist = lsp_to_dist(landform_cove,
#                             dist_fun = "jensen-shannon")
#
# landform_hclust = hclust(landform_dist, method = "ward.D2")
# plot(landform_hclust)
#
# clusters = cutree(landform_hclust, k = 6)
#
# landform_grid_sf = lsp_add_clusters(landform_cove, clusters)
# plot(landform_grid_sf["clust"])
#
# landform_grid_sfq = lsp_add_quality(landform_grid_sf,
#                                        landform_dist)
# plot(landform_grid_sfq["quality"])
library(stars)
library(sf)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
landform_cove = lsp_signature(landform,
                               type = "cove",
                               window = 200,
                               normalization = "pdf")

landform_dist = lsp_to_dist(landform_cove,
                            dist_fun = "jensen-shannon")

landform_hclust = hclust(landform_dist, method = "ward.D2")
#plot(landform_hclust)

clusters = cutree(landform_hclust, k = 4)

landform_grid_sf = lsp_add_clusters(landform_cove, clusters)
#plot(landform_grid_sf["clust"])

#landform_grid_sfq = lsp_add_quality(landform_grid_sf,
#                                        landform_dist)
#plot(landform_grid_sfq["quality"])

## larger data example
# library(stars)
# library(sf)
# landform = read_stars(system.file("raster/landform.tif", package = "motif"),
#                       proxy = FALSE)
# landform_cove = lsp_signature(landform,
#                                type = "cove",
#                                window = 200,
#                                normalization = "pdf")
#
# landform_dist = lsp_to_dist(landform_cove,
#                             dist_fun = "jensen-shannon")
#
# landform_hclust = hclust(landform_dist, method = "ward.D2")
# plot(landform_hclust)
#
# clusters = cutree(landform_hclust, k = 6)
#
# landform_grid_sf = lsp_add_clusters(landform_cove, clusters)
# plot(landform_grid_sf["clust"])
#
# landform_grid_sfq = lsp_add_quality(landform_grid_sf,
#                                        landform_dist)
# plot(landform_grid_sfq["quality"])

Adds spatial data of each region in an lsp or sf object

Description

Adds spatial data of each region in an lsp or sf object. The output is an lsp or sf object with an additional column "region". See examples.

Usage

lsp_add_examples(x, y, window = NULL)

## S3 method for class 'lsp'
lsp_add_examples(x, y, window = NULL)

## S3 method for class 'sf'
lsp_add_examples(x, y, window = NULL)
lsp_add_examples(x, y, window = NULL)

## S3 method for class 'lsp'
lsp_add_examples(x, y, window = NULL)

## S3 method for class 'sf'
lsp_add_examples(x, y, window = NULL)

Arguments

`x`	Object of class `lsp` - usually a subset of the output of `lsp_signature()` or an object of class `sf` - usually a subset of the output of `lsp_search()`
`y`	Object of class `stars`, `stars_proxy`, or terra's `SpatRaster`.
`window`	Specifies areas for analysis. It can be either: `NULL` or an `sf` object. The `sf` object is only needed for adding examples of irregular regions.

Value

The input object with a new column "region". The "region" column is a list with a raster extracted for each row.

Examples


library(stars)

landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))

landcover_coma = lsp_signature(landcover, type = "coma", threshold = 0.9, window = 100)
selected_coma = subset(landcover_coma, id %in% c(5, 10, 15, 35))
selected_coma

selected_coma = lsp_add_examples(x = selected_coma, y = landcover)
selected_coma

plot(selected_coma$region[[1]])
plot(selected_coma$region[[4]])


# larger data example
library(stars)

landcover = read_stars(system.file("raster/landcover2015.tif", package = "motif"))

landcover_coma = lsp_signature(landcover, type = "coma", threshold = 0.9, window = 100)
selected_coma = subset(landcover_coma, id %in% c(5, 80, 1971, 2048))
selected_coma

selected_coma = lsp_add_examples(x = selected_coma, y = landcover)
selected_coma

plot(selected_coma$region[[1]])
plot(selected_coma$region[[4]])

library(stars)

landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))

landcover_coma = lsp_signature(landcover, type = "coma", threshold = 0.9, window = 100)
selected_coma = subset(landcover_coma, id %in% c(5, 10, 15, 35))
selected_coma

selected_coma = lsp_add_examples(x = selected_coma, y = landcover)
selected_coma

plot(selected_coma$region[[1]])
plot(selected_coma$region[[4]])


# larger data example
library(stars)

landcover = read_stars(system.file("raster/landcover2015.tif", package = "motif"))

landcover_coma = lsp_signature(landcover, type = "coma", threshold = 0.9, window = 100)
selected_coma = subset(landcover_coma, id %in% c(5, 80, 1971, 2048))
selected_coma

selected_coma = lsp_add_examples(x = selected_coma, y = landcover)
selected_coma

plot(selected_coma$region[[1]])
plot(selected_coma$region[[4]])

Calculates quality metrics of clustering or segmentation

Description

Calculates three metrics to evaluate quality of spatial patterns' clustering or segmentation. When the type is "cluster", then metrics of inhomogeneity, distinction, and quality are calculated. When the type is "segmentation", then metrics of inhomogeneity, isolation, and quality are calculated. For more information, see Details below.

Usage

lsp_add_quality(x, x_dist, type = "cluster", regions = FALSE)
lsp_add_quality(x, x_dist, type = "cluster", regions = FALSE)

Arguments

`x`	Object of class `sf` - usually the output of the `lsp_add_clusters()` function
`x_dist`	Object of class `dist` - usually the output of the `lsp_to_dist()` function
`type`	Either `"cluster"` or `"segmentation"`
`regions`	Not implemented yet

Details

For type "cluster", this function calculates three quality metrics to evaluate spatial patterns' clustering: (1) inhomogeneity - it measures a degree of mutual dissimilarity between all objects in a cluster. This value is between 0 and 1, where small value indicates that all objects in the cluster represent consistent patterns so the cluster is pattern-homogeneous. (2) distinction - it is an average distance between the focus cluster and all of the other clusters. This value is between 0 and 1, where large value indicates that the cluster stands out from the other clusters. (3) quality - overall quality of a cluster. It is calculated as 1 - (inhomogeneity / distinction). This value is also between 0 and 1, where increased values indicate increased quality.

For type "segmentation", this function calculates three quality metrics to evaluate spatial patterns' segmentation: (1) inhomogeneity - it measures a degree of mutual dissimilarity between all objects in a cluster. This value is between 0 and 1, where small value indicates that all objects in the cluster represent consistent patterns so the cluster is pattern-homogeneous. (2) isolation - it is an average distance between the focus cluster and all of its neighbors. This value is between 0 and 1, where large value indicates that the cluster stands out from its surroundings. (3) quality - overall quality of a cluster. It is calculated as 1 - (inhomogeneity / distinction). This value is also between 0 and 1, where increased values indicate increased quality.

Value

Object of class sf with three additional columns representing quality metrics.

References

Jakub Nowosad & Tomasz F. Stepinski (2021) Pattern-based identification and mapping of landscape types using multi-thematic data, International Journal of Geographical Information Science, DOI: 10.1080/13658816.2021.1893324

Examples

# see examples of `lsp_add_clusters()`

# see examples of `lsp_add_clusters()`

Creates or adds a sf object

Description

Creates or adds a sf object based on the input object or a set of parameters. It accepts either an object of class stars or lsp. In the first case, the output is created based on a set of parameters (window_size and window_shift or window). In the second case, the output converts the lsp object into a sf object.

Usage

lsp_add_sf(x = NULL, window = NULL, metadata = TRUE)

## Default S3 method:
lsp_add_sf(x = NULL, window = NULL, metadata = TRUE)

## S3 method for class 'lsp'
lsp_add_sf(x = NULL, window = NULL, metadata = TRUE)
lsp_add_sf(x = NULL, window = NULL, metadata = TRUE)

## Default S3 method:
lsp_add_sf(x = NULL, window = NULL, metadata = TRUE)

## S3 method for class 'lsp'
lsp_add_sf(x = NULL, window = NULL, metadata = TRUE)

Arguments

`x`	Object of class `stars` or `lsp`. For `stars`, `window` or `window_size` can be used.
`window`	Specifies areas for analysis. It can be either: `NULL`, a numeric value, or an `sf` object. If `window=NULL` calculations are performed for a whole area. If the `window` argument is numeric, it is a length of the side of a square-shaped block of cells. Expressed in the numbers of cells, it defines the extent of a local pattern. If an `sf` object is provided, each feature (row) defines the extent of a local pattern. The `sf` object should have one attribute (otherwise, the first attribute is used as an id).
`metadata`	Logical. Only when ⁠x`` is of class ⁠lsp`⁠. If ⁠`TRUE`⁠, the output object will have metadata ("id" and "na_prop"). If ⁠`FALSE', the output object will not have metadata ("id" and "na_prop").

Value

An sf object converted from the input object or a provided set of parameters

Examples

library(stars)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
plot(landform)
landform_lsp = lsp_add_sf(landform, window = 100)
plot(landform_lsp)

lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
lc_cove_lsp = lsp_add_sf(lc_cove)
plot(lc_cove_lsp["id"])
plot(lc_cove_lsp["na_prop"])


# larger data example
library(stars)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
plot(landform)
landform_lsp = lsp_add_sf(landform, window = 100)
plot(landform_lsp)

lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
lc_cove_lsp = lsp_add_sf(lc_cove)
plot(lc_cove_lsp["id"])
plot(lc_cove_lsp["na_prop"])

library(stars)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
plot(landform)
landform_lsp = lsp_add_sf(landform, window = 100)
plot(landform_lsp)

lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
lc_cove_lsp = lsp_add_sf(lc_cove)
plot(lc_cove_lsp["id"])
plot(lc_cove_lsp["na_prop"])


# larger data example
library(stars)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
plot(landform)
landform_lsp = lsp_add_sf(landform, window = 100)
plot(landform_lsp)

lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
lc_cove_lsp = lsp_add_sf(lc_cove)
plot(lc_cove_lsp["id"])
plot(lc_cove_lsp["na_prop"])

Creates or adds a stars object

Description

Creates or adds a stars object based on the input object or a set of parameters. It accepts either an object of class stars or lsp. In the first case, the output is created based on the window parameter. In the second case, the output converts the lsp object into a stars object.

Usage

lsp_add_stars(x = NULL, window = NULL, metadata = TRUE)

## Default S3 method:
lsp_add_stars(x = NULL, window = NULL, metadata = TRUE)

## S3 method for class 'lsp'
lsp_add_stars(x = NULL, window = NULL, metadata = TRUE)
lsp_add_stars(x = NULL, window = NULL, metadata = TRUE)

## Default S3 method:
lsp_add_stars(x = NULL, window = NULL, metadata = TRUE)

## S3 method for class 'lsp'
lsp_add_stars(x = NULL, window = NULL, metadata = TRUE)

Arguments

`x`	Object of class `stars` or `lsp`. For `stars`, `window` or `window_size` can be used.
`window`	Specifies areas for analysis. It can be either: `NULL`, a numeric value, or an `sf` object. If `window=NULL` calculations are performed for a whole area. If the `window` argument is numeric, it is a length of the side of a square-shaped block of cells. Expressed in the numbers of cells, it defines the extent of a local pattern. If an `sf` object is provided, each feature (row) defines the extent of a local pattern. The `sf` object should have one attribute (otherwise, the first attribute is used as an id).
`metadata`	Logical. Only when ⁠x`` is of class ⁠lsp`⁠. If ⁠`TRUE`⁠, the output object will have metadata ("id" and "na_prop"). If ⁠`FALSE', the output object will not have metadata ("id" and "na_prop").

Value

A stars object converted from the input object or a provided set of parameters

Examples

library(stars)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
plot(landform)
landform_lsp = lsp_add_stars(landform, window = 100)
plot(landform_lsp)

lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
lc_cove_lsp = lsp_add_stars(lc_cove)
plot(lc_cove_lsp)
plot(lc_cove_lsp["na_prop"])



# larger data example
library(stars)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
plot(landform)
landform_lsp = lsp_add_stars(landform, window = 100)
plot(landform_lsp)

lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
lc_cove_lsp = lsp_add_stars(lc_cove)
plot(lc_cove_lsp)
plot(lc_cove_lsp["na_prop"])

library(stars)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
plot(landform)
landform_lsp = lsp_add_stars(landform, window = 100)
plot(landform_lsp)

lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
lc_cove_lsp = lsp_add_stars(lc_cove)
plot(lc_cove_lsp)
plot(lc_cove_lsp["na_prop"])



# larger data example
library(stars)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
plot(landform)
landform_lsp = lsp_add_stars(landform, window = 100)
plot(landform_lsp)

lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
lc_cove_lsp = lsp_add_stars(lc_cove)
plot(lc_cove_lsp)
plot(lc_cove_lsp["na_prop"])

Creates or adds a terra object

Description

Creates or adds a terra object based on the input object or a set of parameters. It accepts either an object of class stars or lsp. In the first case, the output is created based on the window parameter. In the second case, the output converts the lsp object into a terra object.

Usage

lsp_add_terra(x = NULL, window = NULL, metadata = TRUE)
lsp_add_terra(x = NULL, window = NULL, metadata = TRUE)

Arguments

`x`	Object of class `stars` or `lsp`. For `stars`, `window` or `window_size` can be used.
`window`	Specifies areas for analysis. It can be either: `NULL`, a numeric value, or an `sf` object. If `window=NULL` calculations are performed for a whole area. If the `window` argument is numeric, it is a length of the side of a square-shaped block of cells. Expressed in the numbers of cells, it defines the extent of a local pattern. If an `sf` object is provided, each feature (row) defines the extent of a local pattern. The `sf` object should have one attribute (otherwise, the first attribute is used as an id).
`metadata`	Logical. Only when ⁠x`` is of class ⁠lsp`⁠. If ⁠`TRUE`⁠, the output object will have metadata ("id" and "na_prop"). If ⁠`FALSE', the output object will not have metadata ("id" and "na_prop").

Value

A terra object converted from the input object or a provided set of parameters

Examples

library(stars)
library(terra)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
#plot(landform)
landform_lsp = lsp_add_terra(landform, window = 100)
#plot(landform_lsp)

#lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
#lc_cove_lsp = lsp_add_terra(lc_cove)
#plot(lc_cove_lsp)
#plot(lc_cove_lsp["na_prop"])

library(stars)
library(terra)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
#plot(landform)
landform_lsp = lsp_add_terra(landform, window = 100)
#plot(landform_lsp)

#lc_cove = lsp_signature(landform, type = "cove", window = 200, normalization = "pdf")
#lc_cove_lsp = lsp_add_terra(lc_cove)
#plot(lc_cove_lsp)
#plot(lc_cove_lsp["na_prop"])

Comparison between spatial patterns

Description

Compares two spatial datasets containing categorical raster data. It accepts a categorical raster dataset with one or more attributes, and compares it to the second dataset with the same attributes and dimensions. The both dataset are either compared to as whole areas, areas divided into regular windows, or areas divided into irregular windows. This function allows for several types of comparisons using different representations of spatial patterns, including "coma" (co-occurrence matrix), "cove" (co-occurrence vector), "cocoma" (co-located co-occurrence matrix), "cocove" (co-located co-occurrence vector), "wecoma" (weighted co-occurrence matrix), "wecove" (weighted co-occurrence vector), "incoma" (integrated co-occurrence matrix), "incove" (integrated co-occurrence vector). These representations are created for both datasets, and next a distance between them is calculated using a selected measure from the philentropy::distance function. Additional parameters, such as neighbourhood or normalization types, are also available.

Usage

lsp_compare(
  x,
  y,
  type,
  dist_fun,
  window = NULL,
  output = "stars",
  neighbourhood = 4,
  threshold = 0.5,
  ordered = FALSE,
  repeated = FALSE,
  normalization = "pdf",
  wecoma_fun = "mean",
  wecoma_na_action = "replace",
  ...
)
lsp_compare(
  x,
  y,
  type,
  dist_fun,
  window = NULL,
  output = "stars",
  neighbourhood = 4,
  threshold = 0.5,
  ordered = FALSE,
  repeated = FALSE,
  normalization = "pdf",
  wecoma_fun = "mean",
  wecoma_na_action = "replace",
  ...
)

Arguments

`x`	Object of class `stars`, `stars_proxy`, or terra's `SpatRaster`. It should have one attribute (for `"coma"`, `"cove"`), two attributes (`"cocoma"`, `"cocove"`, `"wecoma"`, `"wecove"`), two or more attributes (`"incoma"`, `"incove"`), or any number of attributes suitable for user-defined functions.
`y`	Object of class `stars`, `stars_proxy`, or terra's `SpatRaster`. It should have one attribute (for `"coma"`, `"cove"`), two attributes (`"cocoma"`, `"cocove"`, `"wecoma"`, `"wecove"`), two or more attributes (`"incoma"`, `"incove"`), or any number of attributes suitable for user-defined functions.
`type`	Type of the calculated signature. It can be `"coma"` (co-occurrence matrix), `"cove"` (co-occurrence vector), `"cocoma"` (co-located co-occurrence matrix), `"cocove"` (co-located co-occurrence vector), `"wecoma"` (weighted co-occurrence matrix), `"wecove"` (weighted co-occurrence vector), `"incoma"` (integrated co-occurrence matrix), `"incove"` (integrated co-occurrence vector), `"composition"` or any function that can summarize `stars` objects.
`dist_fun`	Distance measure used. This function uses the `philentropy::distance` function in the background. Run `philentropy::getDistMethods()` to find possible distance measures.
`window`	Specifies areas for analysis. It can be either: `NULL`, a numeric value, or an `sf` object. If `window=NULL` calculations are performed for a whole area. If the `window` argument is numeric, it is a length of the side of a square-shaped block of cells. Expressed in the numbers of cells, it defines the extent of a local pattern. If an `sf` object is provided, each feature (row) defines the extent of a local pattern. The `sf` object should have one attribute (otherwise, the first attribute is used as an id).
`output`	The class of the output. Either `"stars"`, `"sf"`, or `"terra"`
`neighbourhood`	The number of directions in which cell adjacencies are considered as neighbours: 4 (rook's case) or 8 (queen's case). The default is 4.
`threshold`	The share of NA cells to allow metrics calculation.
`ordered`	For `"cove"`, `"cocove"`, `"wecove"` and `"incove"` only. The type of pairs considered. Either "ordered" (TRUE) or "unordered" (FALSE). The default is FALSE.
`repeated`	For `"incove"` only. Should the repeated co-located co-occurrence matrices be used? Either "ordered" (TRUE) or "unordered" (FALSE). The default is FALSE.
`normalization`	For `"cove"`, `"cocove"`, `"wecove"`, `"incove"`, `"composition"`, or user-provided functions only. Should the output vector be normalized? Either "none" or "pdf". The "pdf" option normalizes a vector to sum to one. The default is "pdf".
`wecoma_fun`	For `"wecoma"` and `"wecove"` only. Function to calculate values from adjacent cells to contribute to exposure matrix, `"mean"` - calculate average values of local population densities from adjacent cells, `"geometric_mean"` - calculate geometric mean values of local population densities from adjacent cells, or `"focal"` assign a value from the focal cell
`wecoma_na_action`	For `"wecoma"` and `"wecove"` only. Decides on how to behave in the presence of missing values in `w`. Possible options are `"replace"`, `"omit"`, `"keep"`. The default, `"replace"`, replaces missing values with 0, `"omit"` does not use cells with missing values, and `"keep"` keeps missing values.
`...`	Additional arguments for the `philentropy::distance` function.

Value

Object of class stars (or sf or terra's SpatRaster, depending on the output argument). It has four attributes: (1) id - an id of each window. For irregular windows, it is the values provided in the window argument, (2) na_prop_x - share (0-1) of NA cells for each window in the x object, (3) na_prop_y - share (0-1) of NA cells for each window in the y object, (4) dist- calculated distance between signatures for each window

Examples

library(stars)

lc15 = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))
lc01 = read_stars(system.file("raster/landcover2001s.tif", package = "motif"))
ecoregions = read_sf(system.file("vector/ecoregionss.gpkg", package = "motif"))

ecoregions = st_transform(ecoregions, st_crs(lc15))

c1 = lsp_compare(lc01, lc15, type = "cove",
    dist_fun = "jensen-shannon", window = ecoregions["id"])
plot(c1["dist"])


# larger data example
library(stars)

lc15 = read_stars(system.file("raster/landcover2015.tif", package = "motif"))
lc01 = read_stars(system.file("raster/landcover2001.tif", package = "motif"))
ecoregions = read_sf(system.file("vector/ecoregions.gpkg", package = "motif"))

ecoregions = st_transform(ecoregions, st_crs(lc15))

c1 = lsp_compare(lc01, lc15, type = "cove",
    dist_fun = "jensen-shannon", window = ecoregions["id"])
plot(c1["dist"])

library(stars)

lc15 = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))
lc01 = read_stars(system.file("raster/landcover2001s.tif", package = "motif"))
ecoregions = read_sf(system.file("vector/ecoregionss.gpkg", package = "motif"))

ecoregions = st_transform(ecoregions, st_crs(lc15))

c1 = lsp_compare(lc01, lc15, type = "cove",
    dist_fun = "jensen-shannon", window = ecoregions["id"])
plot(c1["dist"])


# larger data example
library(stars)

lc15 = read_stars(system.file("raster/landcover2015.tif", package = "motif"))
lc01 = read_stars(system.file("raster/landcover2001.tif", package = "motif"))
ecoregions = read_sf(system.file("vector/ecoregions.gpkg", package = "motif"))

ecoregions = st_transform(ecoregions, st_crs(lc15))

c1 = lsp_compare(lc01, lc15, type = "cove",
    dist_fun = "jensen-shannon", window = ecoregions["id"])
plot(c1["dist"])

Extracts a local landscape

Description

Extracts a local landscape from categorical raster data based on its id and provided window argument.

Usage

lsp_extract(x, window, id)
lsp_extract(x, window, id)

Arguments

`x`	Object of class `stars`, `stars_proxy`, or terra's `SpatRaster`.
`window`	Specifies areas for analysis. It can be either: `NULL`, a numeric value, or an `sf` object.
`id`	Id of the local landscape - it is possible to find in the output of `lsp_signature()`, `lsp_search()`, `lsp_compare()`, or `lsp_add_clusters()`.

Value

A starsor terra object cropped to the extent of a selected local landscape

Examples

library(stars)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
ecoregions = read_sf(system.file("vector/ecoregionss.gpkg", package = "motif"))

extract1 = lsp_extract(x = landform, window = 100, id = 25)
plot(extract1)

ecoregions = st_transform(ecoregions, st_crs(landform))
extract2 = lsp_extract(x = landform, window = ecoregions["id"], id = 11)
plot(extract2)


# larger data example
library(stars)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
ecoregions = read_sf(system.file("vector/ecoregions.gpkg", package = "motif"))

extract1 = lsp_extract(x = landform, window = 100, id = 1895)
plot(extract1)

ecoregions = st_transform(ecoregions, st_crs(landform))
extract2 = lsp_extract(x = landform, window = ecoregions["id"], id = 7)
plot(extract2)

library(stars)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
ecoregions = read_sf(system.file("vector/ecoregionss.gpkg", package = "motif"))

extract1 = lsp_extract(x = landform, window = 100, id = 25)
plot(extract1)

ecoregions = st_transform(ecoregions, st_crs(landform))
extract2 = lsp_extract(x = landform, window = ecoregions["id"], id = 11)
plot(extract2)


# larger data example
library(stars)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
ecoregions = read_sf(system.file("vector/ecoregions.gpkg", package = "motif"))

extract1 = lsp_extract(x = landform, window = 100, id = 1895)
plot(extract1)

ecoregions = st_transform(ecoregions, st_crs(landform))
extract2 = lsp_extract(x = landform, window = ecoregions["id"], id = 7)
plot(extract2)

Creates a raster mosaic

Description

Creates a raster mosaic by rearranging spatial data for example regions. See examples.

Usage

lsp_mosaic(x, output = "stars")
lsp_mosaic(x, output = "stars")

Arguments

`x`	Usually the output of the `lsp_add_examples()` function
`output`	The class of the output. Either `"stars"` or `terra`

Value

A stars or terra object

Examples


# larger data example
library(stars)
library(sf)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
landform_cove = lsp_signature(landform,
                               type = "cove",
                               window = 200,
                               normalization = "pdf")

landform_dist = lsp_to_dist(landform_cove,
                            dist_fun = "jensen-shannon")

landform_hclust = hclust(landform_dist, method = "ward.D2")
plot(landform_hclust)

clusters = cutree(landform_hclust, k = 6)

landform_grid_sf = lsp_add_clusters(landform_cove, clusters)
plot(landform_grid_sf["clust"])

landform_grid_sf_sel = landform_grid_sf %>%
    dplyr::filter(na_prop == 0) %>%
    dplyr::group_by(clust) %>%
    dplyr::slice_sample(n = 16, replace = TRUE)

landform_grid_sf_sel = lsp_add_examples(x = landform_grid_sf_sel, y = landform)
landform_grid_sf_sel

landform_clust_m = lsp_mosaic(landform_grid_sf_sel)

plot(landform_clust_m)

# larger data example
library(stars)
library(sf)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
landform_cove = lsp_signature(landform,
                               type = "cove",
                               window = 200,
                               normalization = "pdf")

landform_dist = lsp_to_dist(landform_cove,
                            dist_fun = "jensen-shannon")

landform_hclust = hclust(landform_dist, method = "ward.D2")
plot(landform_hclust)

clusters = cutree(landform_hclust, k = 6)

landform_grid_sf = lsp_add_clusters(landform_cove, clusters)
plot(landform_grid_sf["clust"])

landform_grid_sf_sel = landform_grid_sf %>%
    dplyr::filter(na_prop == 0) %>%
    dplyr::group_by(clust) %>%
    dplyr::slice_sample(n = 16, replace = TRUE)

landform_grid_sf_sel = lsp_add_examples(x = landform_grid_sf_sel, y = landform)
landform_grid_sf_sel

landform_clust_m = lsp_mosaic(landform_grid_sf_sel)

plot(landform_clust_m)

Changes structure of the lsp object

Description

Converts a list-column with signatures into many numeric columns

Usage

lsp_restructure(x)
lsp_restructure(x)

Arguments

`x`	an lsp object

Value

Object of class lsp. It has several columns: (1) id - an id of each window. For irregular windows, it is the values provided in the window argument, (2) na_prop - share (0-1) of NA cells for each window, (3) one or more columns representing values of the signature

Examples


library(stars)

landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))

landcover_cove = lsp_signature(landcover, type = "cove", threshold = 0.9, window = 100)
landcover_cover = lsp_restructure(landcover_cove)
landcover_cover

lsp_add_sf(landcover_cover)
library(stars)

landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))

landcover_cove = lsp_signature(landcover, type = "cove", threshold = 0.9, window = 100)
landcover_cover = lsp_restructure(landcover_cove)
landcover_cover

lsp_add_sf(landcover_cover)

Search for similar spatial pattern

Description

Searches for areas with similar spatial patterns in categorical data. It accepts a categorical raster dataset with one or more attributes, and compares it to the second (usually larger) dataset with the same attributes. The first dataset is either compared to a whole area, areas divided into regular windows, or areas divided into irregular windows from the second dataset. This function allows for several types of comparisons using different representations of spatial patterns, including "coma" (co-occurrence matrix), "cove" (co-occurrence vector), "cocoma" (co-located co-occurrence matrix), "cocove" (co-located co-occurrence vector), "wecoma" (weighted co-occurrence matrix), "wecove" (weighted co-occurrence vector), "incoma" (integrated co-occurrence matrix), "incove" (integrated co-occurrence vector). These representations are created for both datasets, and next a distance between them is calculated using a selected measure from the philentropy::distance function. Additional parameters, such as neighbourhood or normalization types, are also available.

Usage

lsp_search(
  x,
  y,
  type,
  dist_fun,
  window = NULL,
  output = "stars",
  neighbourhood = 4,
  threshold = 0.5,
  ordered = FALSE,
  repeated = FALSE,
  normalization = "pdf",
  wecoma_fun = "mean",
  wecoma_na_action = "replace",
  classes = NULL,
  ...
)
lsp_search(
  x,
  y,
  type,
  dist_fun,
  window = NULL,
  output = "stars",
  neighbourhood = 4,
  threshold = 0.5,
  ordered = FALSE,
  repeated = FALSE,
  normalization = "pdf",
  wecoma_fun = "mean",
  wecoma_na_action = "replace",
  classes = NULL,
  ...
)

Arguments

`x`	Object of class `stars`, `stars_proxy`, or terra's `SpatRaster`. It should have one attribute (for `"coma"`, `"cove"`), two attributes (`"cocoma"`, `"cocove"`, `"wecoma"`, `"wecove"`), two or more attributes (`"incoma"`, `"incove"`), or any number of attributes suitable for user-defined functions.
`y`	Object of class `stars`, `stars_proxy`, or terra's `SpatRaster`. It should have one attribute (for `"coma"`, `"cove"`), two attributes (`"cocoma"`, `"cocove"`, `"wecoma"`, `"wecove"`), two or more attributes (`"incoma"`, `"incove"`), or any number of attributes suitable for user-defined functions.
`type`	Type of the calculated signature. It can be `"coma"` (co-occurrence matrix), `"cove"` (co-occurrence vector), `"cocoma"` (co-located co-occurrence matrix), `"cocove"` (co-located co-occurrence vector), `"wecoma"` (weighted co-occurrence matrix), `"wecove"` (weighted co-occurrence vector), `"incoma"` (integrated co-occurrence matrix), `"incove"` (integrated co-occurrence vector), `"composition"` or any function that can summarize `stars` objects.
`dist_fun`	Distance measure used. This function uses the `philentropy::distance` function in the background. Run `philentropy::getDistMethods()` to find possible distance measures.
`window`	Specifies areas for analysis. It can be either: `NULL`, a numeric value, or an `sf` object. If `window=NULL` calculations are performed for a whole area. If the `window` argument is numeric, it is a length of the side of a square-shaped block of cells. Expressed in the numbers of cells, it defines the extent of a local pattern. If an `sf` object is provided, each feature (row) defines the extent of a local pattern. The `sf` object should have one attribute (otherwise, the first attribute is used as an id).
`output`	The class of the output. Either `"stars"`, `"sf"`, or `"terra"`
`neighbourhood`	The number of directions in which cell adjacencies are considered as neighbours: 4 (rook's case) or 8 (queen's case). The default is 4.
`threshold`	The share of NA cells to allow metrics calculation.
`ordered`	For `"cove"`, `"cocove"`, `"wecove"` and `"incove"` only. The type of pairs considered. Either "ordered" (TRUE) or "unordered" (FALSE). The default is FALSE.
`repeated`	For `"incove"` only. Should the repeated co-located co-occurrence matrices be used? Either "ordered" (TRUE) or "unordered" (FALSE). The default is FALSE.
`normalization`	For `"cove"`, `"cocove"`, `"wecove"`, `"incove"`, `"composition"`, or user-provided functions only. Should the output vector be normalized? Either "none" or "pdf". The "pdf" option normalizes a vector to sum to one. The default is "pdf".
`wecoma_fun`	For `"wecoma"` and `"wecove"` only. Function to calculate values from adjacent cells to contribute to exposure matrix, `"mean"` - calculate average values of local population densities from adjacent cells, `"geometric_mean"` - calculate geometric mean values of local population densities from adjacent cells, or `"focal"` assign a value from the focal cell
`wecoma_na_action`	For `"wecoma"` and `"wecove"` only. Decides on how to behave in the presence of missing values in `w`. Possible options are `"replace"`, `"omit"`, `"keep"`. The default, `"replace"`, replaces missing values with 0, `"omit"` does not use cells with missing values, and `"keep"` keeps missing values.
`classes`	Which classes (categories) should be analyzed? This parameter expects a list of the same length as the number of attributes in `x`, where each element of the list contains integer vector. The default is `NULL`, which means that the classes are calculated directly from the input data and all of them are used in the calculations.
`...`	Additional arguments for the `philentropy::distance` function.

Value

Object of class stars (or sf or terra's SpatRaster, depending on the output argument). It has three attributes: (1) id - an id of each window. For irregular windows, it is the values provided in the window argument, (2) na_prop - share (0-1) of NA cells for each window in the y object, (3) dist- calculated distance between the x object and each window in the y object

Examples

library(stars)

landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))
plot(landcover)

ext = st_bbox(c(xmin = -249797.344531127, xmax = -211162.693944285,
                ymin = -597280.143035389, ymax = -558645.492448547),
                crs = st_crs(landcover))

landcover_ext = landcover[ext]
plot(landcover_ext)

ecoregions = read_sf(system.file("vector/ecoregionss.gpkg", package = "motif"))
plot(ecoregions["id"])

s1 = lsp_search(landcover_ext, landcover, type = "cove",
  dist_fun = "jensen-shannon", threshold = 0.9, window = 100)
plot(s1["dist"])

ecoregions = st_transform(ecoregions, st_crs(landcover))
s2 = lsp_search(landcover_ext, landcover, type = "cove",
  dist_fun = "jensen-shannon", threshold = 0.5, window = ecoregions["id"])
plot(s2["dist"])


# larger data example
library(stars)

landcover = read_stars(system.file("raster/landcover2015.tif", package = "motif"))
plot(landcover)

ext = st_bbox(c(xmin = -249797.344531127, xmax = -211162.693944285,
                ymin = -597280.143035389, ymax = -558645.492448547),
                crs = st_crs(landcover))

landcover_ext = landcover[ext]
plot(landcover_ext)

ecoregions = read_sf(system.file("vector/ecoregions.gpkg", package = "motif"))
plot(ecoregions["id"])

s1 = lsp_search(landcover_ext, landcover, type = "cove",
  dist_fun = "jensen-shannon", threshold = 0.9, window = 1000)
plot(s1["dist"])

ecoregions = st_transform(ecoregions, st_crs(landcover))
s2 = lsp_search(landcover_ext, landcover, type = "cove",
  dist_fun = "jensen-shannon", threshold = 0.5, window = ecoregions["id"])
plot(s2["dist"])

library(stars)

landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))
plot(landcover)

ext = st_bbox(c(xmin = -249797.344531127, xmax = -211162.693944285,
                ymin = -597280.143035389, ymax = -558645.492448547),
                crs = st_crs(landcover))

landcover_ext = landcover[ext]
plot(landcover_ext)

ecoregions = read_sf(system.file("vector/ecoregionss.gpkg", package = "motif"))
plot(ecoregions["id"])

s1 = lsp_search(landcover_ext, landcover, type = "cove",
  dist_fun = "jensen-shannon", threshold = 0.9, window = 100)
plot(s1["dist"])

ecoregions = st_transform(ecoregions, st_crs(landcover))
s2 = lsp_search(landcover_ext, landcover, type = "cove",
  dist_fun = "jensen-shannon", threshold = 0.5, window = ecoregions["id"])
plot(s2["dist"])


# larger data example
library(stars)

landcover = read_stars(system.file("raster/landcover2015.tif", package = "motif"))
plot(landcover)

ext = st_bbox(c(xmin = -249797.344531127, xmax = -211162.693944285,
                ymin = -597280.143035389, ymax = -558645.492448547),
                crs = st_crs(landcover))

landcover_ext = landcover[ext]
plot(landcover_ext)

ecoregions = read_sf(system.file("vector/ecoregions.gpkg", package = "motif"))
plot(ecoregions["id"])

s1 = lsp_search(landcover_ext, landcover, type = "cove",
  dist_fun = "jensen-shannon", threshold = 0.9, window = 1000)
plot(s1["dist"])

ecoregions = st_transform(ecoregions, st_crs(landcover))
s2 = lsp_search(landcover_ext, landcover, type = "cove",
  dist_fun = "jensen-shannon", threshold = 0.5, window = ecoregions["id"])
plot(s2["dist"])

Creates a spatial signature

Description

Calculates selected spatial signatures based on categorical raster data. It also allows for calculations for any defined regular and irregular areas. It has several built-in signatures but also allows for any user-defined functions.

Usage

lsp_signature(
  x,
  type,
  window = NULL,
  neighbourhood = 4,
  threshold = 0.9,
  ordered = FALSE,
  repeated = FALSE,
  normalization = "pdf",
  wecoma_fun = "mean",
  wecoma_na_action = "replace",
  classes = NULL
)
lsp_signature(
  x,
  type,
  window = NULL,
  neighbourhood = 4,
  threshold = 0.9,
  ordered = FALSE,
  repeated = FALSE,
  normalization = "pdf",
  wecoma_fun = "mean",
  wecoma_na_action = "replace",
  classes = NULL
)

Arguments

`x`	Object of class `stars`, `stars_proxy`, or terra's `SpatRaster`. It should have one attribute (for `"coma"`, `"cove"`), two attributes (`"cocoma"`, `"cocove"`, `"wecoma"`, `"wecove"`), two or more attributes (`"incoma"`, `"incove"`), or any number of attributes suitable for user-defined functions.
`type`	Type of the calculated signature. It can be `"coma"` (co-occurrence matrix), `"cove"` (co-occurrence vector), `"cocoma"` (co-located co-occurrence matrix), `"cocove"` (co-located co-occurrence vector), `"wecoma"` (weighted co-occurrence matrix), `"wecove"` (weighted co-occurrence vector), `"incoma"` (integrated co-occurrence matrix), `"incove"` (integrated co-occurrence vector), `"composition"` or any function that can summarize `stars` objects.
`window`	Specifies areas for analysis. It can be either: `NULL`, a numeric value, or an `sf` object. If `window=NULL` calculations are performed for a whole area. If the `window` argument is numeric, it is a length of the side of a square-shaped block of cells. Expressed in the numbers of cells, it defines the extent of a local pattern. If an `sf` object is provided, each feature (row) defines the extent of a local pattern. The `sf` object should have one attribute (otherwise, the first attribute is used as an id).
`neighbourhood`	The number of directions in which cell adjacencies are considered as neighbours: 4 (rook's case) or 8 (queen's case). The default is 4.
`threshold`	The share of NA cells (0-1) to allow metrics calculation.
`ordered`	For `"cove"`, `"cocove"`, `"wecove"` and `"incove"` only. The type of pairs considered. Either "ordered" (TRUE) or "unordered" (FALSE). The default is FALSE.
`repeated`	For `"incove"` only. Should the repeated co-located co-occurrence matrices be used? Either "ordered" (TRUE) or "unordered" (FALSE). The default is FALSE.
`normalization`	For `"cove"`, `"cocove"`, `"wecove"`, `"incove"`, `"composition"`, or user-provided functions only. Should the output vector be normalized? Either "none" or "pdf". The "pdf" option normalizes a vector to sum to one. The default is "pdf".
`wecoma_fun`	For `"wecoma"` and `"wecove"` only. Function to calculate values from adjacent cells to contribute to exposure matrix, `"mean"` - calculate average values of local population densities from adjacent cells, `"geometric_mean"` - calculate geometric mean values of local population densities from adjacent cells, or `"focal"` assign a value from the focal cell
`wecoma_na_action`	For `"wecoma"` and `"wecove"` only. Decides on how to behave in the presence of missing values in `w`. Possible options are `"replace"`, `"omit"`, `"keep"`. The default, `"replace"`, replaces missing values with 0, `"omit"` does not use cells with missing values, and `"keep"` keeps missing values.
`classes`	Which classes (categories) should be analyzed? This parameter expects a list of the same length as the number of attributes in `x`, where each element of the list contains integer vector. The default is `NULL`, which means that the classes are calculated directly from the input data and all of them are used in the calculations.

Value

Object of class lsp. It has three columns: (1) id - an id of each window. For irregular windows, it is the values provided in the window argument, (2) na_prop - share (0-1) of NA cells for each window, (3) signature - a list-column containing calculated signatures

Examples

library(stars)

landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))

landcover_coma = lsp_signature(landcover, type = "coma", threshold = 0.9, window = 2000)
landcover_coma

landcover_comp = lsp_signature(landcover, type = "composition", threshold = 0.9)
landcover_comp


# larger data example
library(stars)

landcover = read_stars(system.file("raster/landcover2015.tif", package = "motif"))

landcover_coma = lsp_signature(landcover, type = "coma", threshold = 0.9, window = 2000)
landcover_coma

landcover_comp = lsp_signature(landcover, type = "composition", threshold = 0.9)
landcover_comp

library(stars)

landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))

landcover_coma = lsp_signature(landcover, type = "coma", threshold = 0.9, window = 2000)
landcover_coma

landcover_comp = lsp_signature(landcover, type = "composition", threshold = 0.9)
landcover_comp


# larger data example
library(stars)

landcover = read_stars(system.file("raster/landcover2015.tif", package = "motif"))

landcover_coma = lsp_signature(landcover, type = "coma", threshold = 0.9, window = 2000)
landcover_coma

landcover_comp = lsp_signature(landcover, type = "composition", threshold = 0.9)
landcover_comp

Calculate Distance Matrix

Description

Calculates a distance matrix based on an object of class lsp.

Usage

lsp_to_dist(x, dist_fun, unit = "log2", p = NULL)
lsp_to_dist(x, dist_fun, unit = "log2", p = NULL)

Arguments

`x`	An object of class `lsp` - usually the output of the `lsp_signature()` function
`dist_fun`	A distance/dissimilarity method used. All possible values can be found using the `philentropy::getDistMethods()` function
`unit`	A character string specifying the logarithm unit that should be used to compute distances that depend on log computations: `⁠"log", "log2", "log10"⁠`. The default is `"log"`
`p`	Power of the Minkowski distance. Used only when the `dist_fun = "minkowski"`

Value

An object of class '"dist"“

Examples

library(stars)
landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))

landcover_cove = lsp_signature(landcover, type = "cove", threshold = 0.9, window = 400)
landcover_cove

dist_cov = lsp_to_dist(landcover_cove, dist_fun = "jensen-shannon")
dist_cov


# larger data example
library(stars)
landcover = read_stars(system.file("raster/landcover2015.tif", package = "motif"))

landcover_cove = lsp_signature(landcover, type = "cove", threshold = 0.9, window = 2000)
landcover_cove

dist_cov = lsp_to_dist(landcover_cove, dist_fun = "jensen-shannon")
dist_cov

library(stars)
landcover = read_stars(system.file("raster/landcover2015s.tif", package = "motif"))

landcover_cove = lsp_signature(landcover, type = "cove", threshold = 0.9, window = 400)
landcover_cove

dist_cov = lsp_to_dist(landcover_cove, dist_fun = "jensen-shannon")
dist_cov


# larger data example
library(stars)
landcover = read_stars(system.file("raster/landcover2015.tif", package = "motif"))

landcover_cove = lsp_signature(landcover, type = "cove", threshold = 0.9, window = 2000)
landcover_cove

dist_cov = lsp_to_dist(landcover_cove, dist_fun = "jensen-shannon")
dist_cov

Transforms lsp objects

Description

It allows for transforming spatial signatures (outputs of the lsp_signature() function) using user-provided functions. See examples for more details.

Usage

lsp_transform(x, fun, ...)
lsp_transform(x, fun, ...)

Arguments

`x`	Object of class `lsp` - usually the output of the `lsp_signature()` function.
`fun`	A user-provided function.
`...`	Additional arguments for `fun`.

Value

Examples

library(stars)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
result_coma500 = lsp_signature(landform, type = "coma", threshold = 0.5, window = 500)

#see how the first signature looks
result_coma500$signature[[1]]

my_function = function(mat){
    mat_c = colSums(mat)
    freqs = mat_c / sum(mat)
    # entropy
    -sum(freqs * log2(freqs), na.rm = TRUE)
}

result_coma500_2 = lsp_transform(result_coma500, my_function)

#see how the first signature looks after transformation
result_coma500_2$signature[[1]]


# larger data example
library(stars)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
result_coma500 = lsp_signature(landform, type = "coma", threshold = 0.5, window = 500)

#see how the first signature looks
result_coma500$signature[[1]]

my_function = function(mat){
    mat_c = colSums(mat)
    freqs = mat_c / sum(mat)
    # entropy
    -sum(freqs * log2(freqs), na.rm = TRUE)
}

result_coma500_2 = lsp_transform(result_coma500, my_function)

#see how the first signature looks after transformation
result_coma500_2$signature[[1]]

library(stars)
landform = read_stars(system.file("raster/landforms.tif", package = "motif"))
result_coma500 = lsp_signature(landform, type = "coma", threshold = 0.5, window = 500)

#see how the first signature looks
result_coma500$signature[[1]]

my_function = function(mat){
    mat_c = colSums(mat)
    freqs = mat_c / sum(mat)
    # entropy
    -sum(freqs * log2(freqs), na.rm = TRUE)
}

result_coma500_2 = lsp_transform(result_coma500, my_function)

#see how the first signature looks after transformation
result_coma500_2$signature[[1]]


# larger data example
library(stars)
landform = read_stars(system.file("raster/landform.tif", package = "motif"))
result_coma500 = lsp_signature(landform, type = "coma", threshold = 0.5, window = 500)

#see how the first signature looks
result_coma500$signature[[1]]

my_function = function(mat){
    mat_c = colSums(mat)
    freqs = mat_c / sum(mat)
    # entropy
    -sum(freqs * log2(freqs), na.rm = TRUE)
}

result_coma500_2 = lsp_transform(result_coma500, my_function)

#see how the first signature looks after transformation
result_coma500_2$signature[[1]]

Prepares window* arguments (internal function)

Description

Prepares window* arguments (internal function)

Usage

prepare_window(x, window)
prepare_window(x, window)

Arguments

x

Object of class stars or stars_proxy

window

Specifies areas for analysis. It can be either: NULL, a numeric value, or an sf object. If window=NULL calculations are performed for a whole area. If the window argument is numeric, it is a length of the side of a square-shaped block of cells. Expressed in the numbers of cells, it defines the extent of a local pattern. If an sf object is provided, each feature (row) defines the extent of a local pattern. The sf object should have one attribute (otherwise, the first attribute is used as an id).

Value

A list with window, window_size, and window_shift

Package 'motif'

Help Index

Determine unique classes (internal function)

Description

Usage

Arguments

Value

Adds clusters' ids to a lsp object

Description

Usage

Arguments

Value

Examples

Adds spatial data of each region in an lsp or sf object

Description

Usage

Arguments

Value

Examples

Calculates quality metrics of clustering or segmentation

Description

Usage

Arguments

Details

Value

References

See Also

Examples

Creates or adds a sf object

Description

Usage

Arguments

Value

Examples

Creates or adds a stars object

Description

Usage

Arguments

Value

Examples

Creates or adds a terra object

Description

Usage

Arguments

Value

Examples

Comparison between spatial patterns

Description

Usage

Arguments

Value

Examples

Extracts a local landscape

Description

Usage

Arguments

Value

Examples

Creates a raster mosaic

Description

Usage

Arguments

Value

Examples

Changes structure of the lsp object

Description

Usage

Arguments

Value

Examples

Search for similar spatial pattern

Description

Usage

Arguments

Value

Examples

Creates a spatial signature

Description

Usage

Arguments