NEWS.md
In JosiahParry/sfdep: Spatial Dependence for Simple Features

sfdep (development version)

Fixes bug with cli and closes #49

sfdep 0.2.4

Dexter Locke is now the maintainer of sfdep
Address failing CRAN checks h/t @rsbivand https://github.com/JosiahParry/sfdep/pull/43
Address bug in local Gi/* statistics

sfdep 0.2.3

remove readr suggestion to pass noSuggests from CRAN. Much minuscule tasks for the overlords.

sfdep 0.2.2

no changes from 0.2.1 version bump solely for CRAN

sfdep 0.2.1

Bivariate Moran uses the new spdep implementation which is faster. It does, however, return an object of class boot which needs extra work to calculate the p-value
Local join count univariate uses spdep now
adds graph based neighbors:
st_nb_delaunay() uses spdep::tri2nb()
st_nb_gabriel() uses spdep::gabrielneigh() and spdep::graph2nb()
st_nb_relative() uses spdep::relativeneigh() and spdep::graph2nb()
activate() is now exported as a generic method to no longer conflict with tidygraph
st_as_graph() now creates undirected graphs
new functions for sfnetworks
node_get_nbs(): converts the node adjacency list to a neighbor list
node_get_edge_list(): creates an edge adjacency list used to get edge attribute columns
node_get_edge_col(): uses the edge list to get edge attributes and turn them into a weight like list

sfdep 0.2.0

spatial_gini() is a new global measure
euclidean_median() is added for identifying the euclidean median of points for point pattern analysis.
Implemented using pracma::geo_median()
- pracma is now added as a suggested package
center_mean() and center_median() are added for point pattern analysis
tally_jc() returns join count matrix via spdep::joincount.multi() as a data.frame object
st_complete_nb() creates a complete graph of neighbors
set operations for neighbor lists nb_union(), nb_intersect(), nb_setdiff()
st_block_nb() is introduced to create neighbor contiguity based on spatial regimes
inspired by https://pysal.org/libpysal/generated/libpysal.weights.block_weights.html
critical_threshold() gains argument k to identify minimum number of neighbors when calculating threshold
pct_nonzero() calculates the percent of non-zero neighbors
szero() to calculate the global sum of weights
nb_as_matrix() and wt_as_matrix() convert neighbor and weights lists to matrices
two new functions include_self() based on spdep::include.self() and new remove_self() which is available separately in spdep due to issue 83 as spdep::remove.self().
local_g_perm() returns correct G statistic. Previously returned only the internal measure.
local_gstar() and local_gstar_perm() do not warn when self.included = TRUE
emerging_hotspot_analysis() implements emerging hotspot analysis with spacetime objects
a new spacetime class for linking spatio-temporal data and geometry
see the spacetime vignette.
Implemented colocation quotient (CLQ) measures See the colocation vignette:
global_colocation() for global colocation of one categorical variable
pairwise_colocation() for asymmetric colocation of subcategories of one or two categories
local_colocation() for local CLQ measure
these functions are not fast! PRs welcomed to improve computation speed.
st_kernel_weights() now returns an attribute kernel which is set to the value of the kernel argument
cond_permute_nb() specifies SIMPLIFY = FALSE in internal mapply() call which was causing errors in creating conditionally permuted neighbor lists