new big realease, still working on details before submiting to CRAN

DESCRIPTION

@@ -47,6 +47,7 @@ Suggests:
     knitr,
     rmarkdown
 RoxygenNote: 7.2.3
+Roxygen: list(markdown = TRUE)
 VignetteBuilder: 
     knitr
 URL: https://jeremygelb.github.io/spNetwork/

NEWS.md

@@ -1,3 +1,15 @@
+# spNetwork 0.4.4
+
+This is an important release!
+
+## New Features
+
+* It is now possible to do bandwidth selection with adaptive bandwidths
+* One can also set local bandwidths manually for bandwidth selection
+
+## Changes in function
+* for all the function where the parameters *bw_range* and *bw_step* were used, they were replaced by a single argument *bws*. It gives more flexibility to the user. 
+
 # spNetwork 0.4.3.7
 
 This is only a maintenance release. A bug in CRAN caused by multiprocessing in example caused the rejection of hte package.

R/adaptive_simultaneous_tnkde_bw.R

@@ -69,11 +69,8 @@ worker_adaptive_bw_tnkde <- function(lines,
   setDT(quad_events2)[events_loc2, on = "goid", vertex_id := i.vertex_id]
 
   ## step3 starting the calculations !
-  print(graph)
   neighbour_list <- adjacent_vertices(graph,nodes$id,mode="out")
   neighbour_list <- lapply(neighbour_list,function(x){return (as.numeric(x))})
-  print("here is the neighbour list : ")
-  print(neighbour_list)
 
   kernel_values <- adaptive_bw_tnkde_cpp2(method = method,
                                          neighbour_list = neighbour_list,
@@ -164,7 +161,7 @@ adaptive_bw_tnkde <- function(grid, events_loc, events, lines,
   n_quadra <- length(selections)
 
   dfs <- lapply(1:n_quadra,function(i){
-    print(i)
+
     sel <- selections[[i]]
 
     # sel_events gives us the events inside the quadra considered
@@ -211,8 +208,6 @@ adaptive_bw_tnkde <- function(grid, events_loc, events, lines,
   final_bws_net <- array(0,dim = dim(tot_arr))
   final_bws_time <- array(0,dim = dim(tot_arr))
 
-  print("here are the estimated densities before calculating local bws")
-  print(tot_arr)
   for(i in 1:length(bw_net)){
     for(j in 1:length(bw_time)){
       k <- tot_arr[i,j,]
@@ -367,8 +362,6 @@ adaptive_bw_tnkde.mc <- function(grid, events_loc, events, lines,
   final_bws_net <- array(0,dim = dim(tot_arr))
   final_bws_time <- array(0,dim = dim(tot_arr))
 
-  print("here are the estimated densities before calculating local bws")
-  print(tot_arr)
   for(i in 1:length(bw_net)){
     for(j in 1:length(bw_time)){
       k <- tot_arr[i,j,]

R/bandwidth_selection_cv_sf.R

@@ -236,10 +236,10 @@
 #' the dataset (leave one out cross validation). We use here the shortcut formula as
 #' described by the package spatstat \insertCite{spatstatpkg}{spNetwork}.
 #'
-#' LCV(h) = sum[i] log(lambda[-i](x[i]))
+#'  \eqn{LCV(h) = \sum_i \log\hat\lambda_{-i}(x_i)}
 #'
-#' Where the sum is taken for all events x[i] and where lambda[-i](x[i]) is the leave-one-out kernel
-#' estimate at x[i] for a bandwidth h. A higher value indicates a better bandwidth.
+#' Where the sum is taken for all events \eqn{x_i} and where \eqn{\hat\lambda_{-i}(x_i)} is the leave-one-out kernel
+#' estimate at \eqn{x_i} for a bandwidth h. A higher value indicates a better bandwidth.
 #'
 #' @references{
 #'     \insertAllCited{}
@@ -299,7 +299,7 @@ bw_cv_likelihood_calc <- function(bws = NULL,
 
   passed <- bw_checks(check,lines,samples,events,
            kernel_name, method, bws_net = bws,
-           adaptive = adaptive, trim_net_bws = trim_bws,
+           adaptive = adaptive, trim_net_bws = trim_bws, arr_bws_net = mat_bws,
            diggle_correction = diggle_correction, study_area = study_area)
 
   if(zero_strat %in% c("min_double", "remove") == FALSE){
@@ -464,7 +464,7 @@ bw_cv_likelihood_calc <- function(bws = NULL,
 #' select an appropriate bandwidth in a data-driven approach
 #'
 #' @details  See the function bw_cv_likelihood_calc for more details. The calculation is split
-#' according to the parameter grid_shape. If grid_shape = c(1,1), then parallel processing cannot be used.
+#' according to the parameter grid_shape. If `grid_shape = c(1,1)`, then parallel processing cannot be used.
 #'
 #' @template bw_selection-args
 #' @template diggle_corr-arg
@@ -522,7 +522,7 @@ bw_cv_likelihood_calc.mc <- function(bws, lines, events, w, kernel_name, method,
 
   passed <- bw_checks(check,lines,samples,events,
                       kernel_name, method, bws_net = bws, bws_time = NULL,
-                      adaptive = adaptive, trim_net_bws = trim_bws,
+                      adaptive = adaptive, trim_net_bws = trim_bws, arr_bws_net = mat_bws,
                       diggle_correction = diggle_correction, study_area = study_area)
 
   if(zero_strat %in% c("min_double", "remove") == FALSE){