mpadge.tngl.sh
Distances on Directed Graphs in R
1# Generated by using Rcpp::compileAttributes() -> do not edit by hand
2# Generator token: 10BE3573-1514-4C36-9D1C-5A225CD40393
3
4#' rcpp_centrality - parallel function
5#'
6#' sample is used to estimate timing, by calculating centrality from just a few
7#' vertices.
8#' @noRd
9rcpp_centrality <- function (graph, vert_map_in, heap_type, dist_threshold, edge_centrality, sample) {
10 .Call (`_dodgr_rcpp_centrality`, graph, vert_map_in, heap_type, dist_threshold, edge_centrality, sample)
11}
12
13#' rcpp_concaveman
14#' @noRd
15rcpp_concaveman <- function (xy, hull_in, concavity, length_threshold) {
16 .Call (`_dodgr_rcpp_concaveman`, xy, hull_in, concavity, length_threshold)
17}
18
19#' De-duplicate edges by replacing with minimal weighted distances and times
20#'
21#' @param graph Full graph in any form
22#' @param fr_col Name of column holding from edge labels
23#' @param to_col Name of column holding to edge labels
24#' @param d_col Name of column holding weighted distances
25#' @param t_col Name of column holding weighted times (or "" if no weighted
26#' times).
27#' @return A `data.frame` of 3 columns: 'from', 'to', and 'd', where 'd' is the
28#' minimal value taken from all duplicated edges. If 't_col' is specified, the
29#' equivalent minimal times are in the lower half of the result.
30#' @noRd
31rcpp_deduplicate <- function (graph, fr_col, to_col, d_col, t_col) {
32 .Call (`_dodgr_rcpp_deduplicate`, graph, fr_col, to_col, d_col, t_col)
33}
34
35#' Make unordered_set of all new edge names
36#' @noRd
37NULL
38
39#' Make vector of all new edge names
40#' @noRd
41NULL
42
43#' Get numbers of old edges corresponding to each contracted edge
44#' @noRd
45NULL
46
47#' Collect sets of all from and to vertices for each set of edges corresponding
48#' to each contracted edge. These sets aren't in any particular order, but the
49#' two sets may be used to match the from and to vertices. These sets are then
50#' arranged into sequences in the subsequent function,
51#' \code{order_vert_sequences}.
52#' @noRd
53NULL
54
55#' rcpp_aggregate_to_sf
56#'
57#' Aggregate a dodgr network data.frame to an sf LINESTRING data.frame
58#'
59#' @param graph_full Rcpp::DataFrame containing the **full** graph
60#' @param graph_contr Rcpp::DataFrame containing the **contracted** graph
61#' @param edge_map Rcpp::DataFrame containing the edge map returned from
62#' \code{dodgr_contract_graph}
63#'
64#' @return Rcpp::List object of `sf::LINESTRING` geoms
65#'
66#' @noRd
67rcpp_aggregate_to_sf <- function (graph_full, graph_contr, edge_map) {
68 .Call (`_dodgr_rcpp_aggregate_to_sf`, graph_full, graph_contr, edge_map)
69}
70
71#' rcpp_flows_aggregate_par
72#'
73#' @param graph The data.frame holding the graph edges
74#' @param vert_map_in map from <std::string> vertex ID to (0-indexed) integer
75#' index of vertices
76#' @param fromi Index into vert_map_in of vertex numbers
77#' @param toi Index into vert_map_in of vertex numbers
78#' @param tol Relative tolerance in terms of flows below which targets
79#' (to-vertices) are not considered.
80#'
81#' @note The parallelisation is achieved by dumping the results of each thread
82#' to a file, with aggregation performed at the end by simply reading back and
83#' aggregating all files. There is no way to aggregate into a single vector
84#' because threads have to be independent. The only danger with this approach
85#' is that multiple threads may generate the same file names, but with names 10
86#' characters long, that chance should be 1 / 62 ^ 10.
87#'
88#' @noRd
89rcpp_flows_aggregate_par <- function (graph, vert_map_in, fromi, toi_in, flows, norm_sums, tol, heap_type) {
90 .Call (`_dodgr_rcpp_flows_aggregate_par`, graph, vert_map_in, fromi, toi_in, flows, norm_sums, tol, heap_type)
91}
92
93#' rcpp_flows_aggregate_pairwise
94#'
95#' Pairwise version of flows_aggregate_par
96#'
97#' @param graph The data.frame holding the graph edges
98#' @param vert_map_in map from <std::string> vertex ID to (0-indexed) integer
99#' index of vertices
100#' @param fromi Index into vert_map_in of vertex numbers
101#' @param toi Index into vert_map_in of vertex numbers
102#' @param tol Relative tolerance in terms of flows below which targets
103#' (to-vertices) are not considered.
104#'
105#' @note The parallelisation is achieved by dumping the results of each thread
106#' to a file, with aggregation performed at the end by simply reading back and
107#' aggregating all files. There is no way to aggregate into a single vector
108#' because threads have to be independent. The only danger with this approach
109#' is that multiple threads may generate the same file names, but with names 10
110#' characters long, that chance should be 1 / 62 ^ 10.
111#'
112#' @noRd
113rcpp_flows_aggregate_pairwise <- function (graph, vert_map_in, fromi, toi, flows, norm_sums, tol, heap_type) {
114 .Call (`_dodgr_rcpp_flows_aggregate_pairwise`, graph, vert_map_in, fromi, toi, flows, norm_sums, tol, heap_type)
115}
116
117#' rcpp_flows_disperse_par
118#'
119#' Modified version of \code{rcpp_flows_aggregate} that aggregates flows to all
120#' destinations from given set of origins, with flows attenuated by distance
121#' from those origins.
122#'
123#' @param graph The data.frame holding the graph edges
124#' @param vert_map_in map from <std::string> vertex ID to (0-indexed) integer
125#' index of vertices
126#' @param fromi Index into vert_map_in of vertex numbers
127#' @param k Coefficient of (current proof-of-principle-only) exponential
128#' distance decay function. If value of \code{k<0} is given, a standard
129#' logistic polynomial will be used.
130#'
131#' @note The flow data to be used for aggregation is a matrix mapping flows
132#' between each pair of from and to points.
133#'
134#' @noRd
135rcpp_flows_disperse_par <- function (graph, vert_map_in, fromi, k, dens, tol, heap_type) {
136 .Call (`_dodgr_rcpp_flows_disperse_par`, graph, vert_map_in, fromi, k, dens, tol, heap_type)
137}
138
139#' rcpp_flows_si
140#'
141#' @param graph The data.frame holding the graph edges
142#' @param vert_map_in map from <std::string> vertex ID to (0-indexed) integer
143#' index of vertices
144#' @param fromi Index into vert_map_in of vertex numbers
145#' @param toi Index into vert_map_in of vertex numbers
146#' @param kvec Vector of k-values for each fromi
147#' @param nvec Vector of density-values for each fromi
148#' @param tol Relative tolerance in terms of flows below which targets
149#' (to-vertices) are not considered.
150#'
151#' @noRd
152rcpp_flows_si <- function (graph, vert_map_in, fromi, toi_in, kvec, dens_from, dens_to, norm_sums, tol, heap_type) {
153 .Call (`_dodgr_rcpp_flows_si`, graph, vert_map_in, fromi, toi_in, kvec, dens_from, dens_to, norm_sums, tol, heap_type)
154}
155
156#' @noRd
157rcpp_fundamental_cycles <- function (graph, verts) {
158 .Call (`_dodgr_rcpp_fundamental_cycles`, graph, verts)
159}
160
161#' get_to_from
162#'
163#' Get one pair of two and from edges and vertices. Main task is to make sure
164#' that bi-directed edges ("intermediate_double") correctly return the
165#' **different** values of from and to vertices and edges.
166#'
167#' @noRd
168NULL
169
170#' same_hwy_type
171#'
172#' Determine whether two edges represent the same weight category (type of
173#' highway for street networks, for example). Categories are not retained in
174#' converted graphs, but can be discerned by comparing ratios of weighted to
175#' non-weighted distances.
176#' @noRd
177NULL
178
179#' rcpp_contract_graph
180#'
181#' Removes nodes and edges from a graph that are not needed for routing
182#'
183#' @param graph graph to be processed
184#'
185#' @return \code{Rcpp::List} containing one \code{data.frame} with the
186#' contracted graph, one \code{data.frame} with the original graph and one
187#' \code{data.frame} containing information about the relating edge ids of the
188#' original and contracted graph.
189#'
190#' @noRd
191rcpp_contract_graph <- function (graph, vertlist_in) {
192 .Call (`_dodgr_rcpp_contract_graph`, graph, vertlist_in)
193}
194
195#' rcpp_merge_cols
196#'
197#' Merge columns in directed graph to form aggregate undirected columns, and
198#' return a corresponding undirected graph useful for visualisation.
199#'
200#' @param graph The result of a call to \code{dodgr_flows_aggregate/disperse}
201#' or similar function resuling in columns of directed values.
202#' @return A single vector of aggregate values with non-zero values only for
203#' those edges to be retained in the directed graph.
204#'
205#' @noRd
206rcpp_merge_cols <- function (graph) {
207 .Call (`_dodgr_rcpp_merge_cols`, graph)
208}
209
210#' sample_one_edge_no_comps
211#'
212#' Sample one edge for graph that has no pre-calculated components. Only used
213#' in \code{sample_one_vertex}
214#'
215#' @param edge_map edge_map
216#' @return std::vector of 2 elements: [0] with value of largest connected
217#' component; [1] with random index to one edge that is part of that component.
218#' @noRd
219NULL
220
221#' sample_one_edge_with_comps
222#'
223#' Sample one edge for graph that has pre-calculated components. Only used in
224#' \code{sample_one_vertex}
225#'
226#' @param edge_map edge_map
227#' @return Random index to one edge that is part of the largest connected
228#' component.
229#' @noRd
230NULL
231
232#' rcpp_sample_graph
233#'
234#' Randomly sample one connected componnent of a graph
235#'
236#' @param graph graph to be processed
237#' @param nverts_to_sample Number of vertices to sample
238#' @param quiet If TRUE, display progress
239#'
240#' @return Smaller sub-set of \code{graph}
241#'
242#' @noRd
243rcpp_sample_graph <- function (graph, nverts_to_sample) {
244 .Call (`_dodgr_rcpp_sample_graph`, graph, nverts_to_sample)
245}
246
247#' graph_has_components
248#'
249#' Does a graph have a vector of connected component IDs? Only used in
250#' \code{sample_one_vertex}
251#' @noRd
252NULL
253
254#' @name graph_from_df
255#'
256#' Convert a standard graph data.frame into an object of class graph. Graphs
257#' are standardised with the function \code{dodgr_convert_graph()$graph}, and
258#' contain only the four columns [from, to, d, w]
259#'
260#' @noRd
261NULL
262
263#' identify_graph_components
264#'
265#' Identify initial graph components for each **vertex**
266#' Identification for edges is subsequently perrformed with
267#' \code{rcpp_get_component_vector}.
268#'
269#' @param v unordered_map <vertex_id_t, vertex_t>
270#' @param com component map from each vertex to component numbers
271#' @noRd
272NULL
273
274#' rcpp_get_component_vector
275#'
276#' Get component numbers for each edge of graph
277#'
278#' @param graph graph to be processed; stripped down and standardised to five
279#' columns
280#'
281#' @return Two vectors: one of edge IDs and one of corresponding component
282#' numbers
283#' @noRd
284rcpp_get_component_vector <- function (graph) {
285 .Call (`_dodgr_rcpp_get_component_vector`, graph)
286}
287
288#' rcpp_unique_rownames
289#'
290#' Construct vertex (from, to) ID values from unique pairs of coordinates
291#' rounded to <precision>. Used when vertices have no ID values.
292#'
293#' @noRd
294rcpp_unique_rownames <- function (xyfrom, xyto, precision = 10L) {
295 .Call (`_dodgr_rcpp_unique_rownames`, xyfrom, xyto, precision)
296}
297
298#' Determine which side of intersecting line a point lies on.
299#'
300#' @param (ax, ay) Coordinates of one end of line
301#' @param (bx, by) Coordinates of other end of line
302#' @param (x, y) Coordinates of point
303#' @return 0 if point on line, -1 if to the left; +1 if to the right.
304#'
305#' @noRd
306NULL
307
308#' Simple match of points to nearest vertices
309#' @noRd
310NULL
311
312#' Match points to nearest edge of graph at which perpendicular from point
313#' bisects edges. Uses psuedo-code from
314#' https://stackoverflow.com/a/6853926
315#' @noRd
316NULL
317
318#' rcpp_points_index_par
319#'
320#' Get index of nearest vertices to list of points
321#'
322#' @param xy Rcpp::DataFrame containing the vertex coordinates of the graph
323#' @param pts Rcpp::DataFrame containing the points to be matched
324#'
325#' @return 0-indexed Rcpp::NumericVector index into graph of nearest points
326#'
327#' @noRd
328rcpp_points_index_par <- function (xy, pts) {
329 .Call (`_dodgr_rcpp_points_index_par`, xy, pts)
330}
331
332#' rcpp_points_to_edges_par
333#'
334#' Get index of nearest edges to list of points
335#'
336#' @param graph Rcpp::DataFrame containing the full edge-based graph
337#' @param pts Rcpp::DataFrame containing the points to be matched
338#'
339#' @return 0-indexed Rcpp::NumericVector index into graph of nearest points
340#'
341#' @noRd
342rcpp_points_to_edges_par <- function (graph, pts) {
343 .Call (`_dodgr_rcpp_points_to_edges_par`, graph, pts)
344}
345
346#' rcpp_get_sp_dists_par
347#'
348#' @noRd
349rcpp_get_sp_dists_par <- function (graph, vert_map_in, fromi, toi_in, heap_type, is_spatial) {
350 .Call (`_dodgr_rcpp_get_sp_dists_par`, graph, vert_map_in, fromi, toi_in, heap_type, is_spatial)
351}
352
353#' rcpp_get_sp_dists_nearest
354#'
355#' @noRd
356rcpp_get_sp_dists_nearest <- function (graph, vert_map_in, fromi, toi_in, heap_type) {
357 .Call (`_dodgr_rcpp_get_sp_dists_nearest`, graph, vert_map_in, fromi, toi_in, heap_type)
358}
359
360#' rcpp_get_sp_dists_paired_par
361#'
362#' @noRd
363rcpp_get_sp_dists_paired_par <- function (graph, vert_map_in, fromi, toi, heap_type, is_spatial) {
364 .Call (`_dodgr_rcpp_get_sp_dists_paired_par`, graph, vert_map_in, fromi, toi, heap_type, is_spatial)
365}
366
367#' rcpp_get_iso
368#'
369#' @noRd
370rcpp_get_iso <- function (graph, vert_map_in, fromi, dlim, heap_type) {
371 .Call (`_dodgr_rcpp_get_iso`, graph, vert_map_in, fromi, dlim, heap_type)
372}
373
374#' rcpp_get_sp_dists
375#'
376#' @noRd
377rcpp_get_sp_dists <- function (graph, vert_map_in, fromi, toi_in, heap_type) {
378 .Call (`_dodgr_rcpp_get_sp_dists`, graph, vert_map_in, fromi, toi_in, heap_type)
379}
380
381#' rcpp_get_paths
382#'
383#' @param graph The data.frame holding the graph edges
384#' @param vert_map_in map from <std::string> vertex ID to (0-indexed) integer
385#' index of vertices
386#' @param fromi Index into vert_map_in of vertex numbers
387#' @param toi Index into vert_map_in of vertex numbers
388#'
389#' @note The graph is constructed with 0-indexed vertex numbers contained in
390#' code{vert_map_in}. Both \code{fromi} and \code{toi} already map directly
391#' onto these. The graph has to be constructed by first constructing a
392#' \code{std::map} object (\code{vertmap}) for \code{vert_map_in}, then
393#' translating all \code{graph["from"/"to"]} values into these indices. This
394#' construction is done in \code{inst_graph}.
395#'
396#' @note Returns 1-indexed values indexing directly into the R input
397#'
398#' @noRd
399rcpp_get_paths <- function (graph, vert_map_in, fromi, toi_in, heap_type) {
400 .Call (`_dodgr_rcpp_get_paths`, graph, vert_map_in, fromi, toi_in, heap_type)
401}
402
403rcpp_get_paths_pairwise <- function (graph, vert_map_in, fromi, toi_in, heap_type) {
404 .Call (`_dodgr_rcpp_get_paths_pairwise`, graph, vert_map_in, fromi, toi_in, heap_type)
405}
406
407#' rcpp_get_sp_dists_categorical
408#'
409#' The `graph` must have an `edge_type` column of non-negative integers,
410#' with 0 denoting edges which are not aggregated, and all other values
411#' defining aggregation categories.
412#'
413#' Implemented in parallal form only; no single-threaded version, and
414#' only for AStar (so graphs must be spatial).
415#' @noRd
416rcpp_get_sp_dists_categorical <- function (graph, vert_map_in, fromi, toi_in, heap_type, proportions_only) {
417 .Call (`_dodgr_rcpp_get_sp_dists_categorical`, graph, vert_map_in, fromi, toi_in, heap_type, proportions_only)
418}
419
420#' rcpp_get_sp_dists_categ_paired
421#'
422#' Pairwise version of 'get_sp_dists_categorical'. The `graph` must have an
423#' `edge_type` column of non-negative integers, with 0 denoting edges which are
424#' not aggregated, and all other values defining aggregation categories.
425#'
426#' Implemented in parallal form only; no single-threaded version, and
427#' only for AStar (so graphs must be spatial).
428#' @noRd
429rcpp_get_sp_dists_categ_paired <- function (graph, vert_map_in, fromi, toi_in, heap_type) {
430 .Call (`_dodgr_rcpp_get_sp_dists_categ_paired`, graph, vert_map_in, fromi, toi_in, heap_type)
431}
432
433#' rcpp_get_sp_dists_cat_threshold
434#'
435#' The `graph` must have an `edge_type` column of non-negative integers,
436#' with 0 denoting edges which are not aggregated, and all other values
437#' defining aggregation categories.
438#'
439#' Implemented in parallal form only; no single-threaded version, and
440#' only for AStar (so graphs must be spatial).
441#' @noRd
442rcpp_get_sp_dists_cat_threshold <- function (graph, vert_map_in, fromi, dlimit, heap_type) {
443 .Call (`_dodgr_rcpp_get_sp_dists_cat_threshold`, graph, vert_map_in, fromi, dlimit, heap_type)
444}
445
446#' rcpp_gen_hash
447#'
448#' Efficient generation of long sequences of hash keys
449#'
450#' @noRd
451rcpp_gen_hash <- function (n, hash_len) {
452 .Call (`_dodgr_rcpp_gen_hash`, n, hash_len)
453}
454
455#' rcpp_sf_as_network
456#'
457#' Return OSM data from Simple Features format input
458#'
459#' @param sf_lines An sf collection of LINESTRING objects
460#' @param pr Rcpp::DataFrame containing the weighting profile
461#'
462#' @return Rcpp::List objects of OSM data, one matrix of numeric and one of
463#' character values. The former contain 7 columns:
464#' 1. sf geom index
465#' 2. from longitude
466#' 3. from latitude
467#' 4. to longitude
468#' 5. to latitude
469#' 6. distance
470#' 7. weighted_distance
471#' The character value matrix has 4 columns of:
472#' 1. from ID
473#' 2. to ID
474#' 3. highway type
475#' 4. OSM way ID
476#'
477#' @noRd
478rcpp_sf_as_network <- function (sf_lines, pr) {
479 .Call (`_dodgr_rcpp_sf_as_network`, sf_lines, pr)
480}
481
482#' rcpp_route_times
483#'
484#' @noRd
485rcpp_route_times <- function (graph, left_side, turn_penalty) {
486 .Call (`_dodgr_rcpp_route_times`, graph, left_side, turn_penalty)
487}