added unweighted networklevel indices to the wishlist (wish by Mark M…

…urphy)

Wishlist.txt

@@ -7,9 +7,10 @@ large
 	** revise plotweb (incl. device settings, vertical option, margin control; easy and explained coloring per sp etc)
 
 * recursive modules with Beckett; also check for how negative values are used (apparently different by the two algorithms); are they locally defined?
+* for each network index, run timing on increasing large networks (up to millions of cells): which indices are available for REALLY large networks? -> add that as index choice!
 
-* NULLMODEL-LIKE:
-* in nullmodel, the check for binary networks is flawed for numbers < 1; use length(unique(web)<3) instead
+* NULLMODEL-LIKE:
+* in nullmodel, the check for binary networks is flawed for numbers < 1; use length(unique(web)<3) instead
 * also, check that values are large enough for r2dtable (small values will be rounded down to 0, making the whole null model approach wrong!)
 * use dirmult (in the same package) to fit Dirichlet-distribution to network data; use these to draw from random distributions in null models?
 * Guillaume & Latapy: implement random bipartite graph algorithm!
@@ -25,6 +26,7 @@ large
 small
 * FIX VEGAN DEPENDENCY!! commsimulator goes out of business, see email by Oksanen 25.4.2018!
 (improve existing functionality)
+* add UNWEIGHTED vulnerability/generality/linkagedensity/weighted connectance to networklevel (see Email by Mark Murphy, 17-Oct-2018)
 * test specieslevel with empty.web=FALSE; decide where web.e must be used
 * better treatment of allowed index synonyms in species/group/networklevel (currently they are "not recognised" if they don't come with a allindex-namevariant)
 * make spaces in names consistent: input(=index) should be the same as output (in list)
@@ -39,6 +41,7 @@ small
 * include robustnessLD50 (Dunne et al. 2002, MacFadyen et al. 2009) and revise help for robustness accordingly
 * search for JFedit; then cleanup comments in species/group/networklevel .R / .Rd
 * functions accepting external abundances (esp. plotweb); check what happens with unnamed abun-vectors and maybe give warnings
+* clean up authors and add ORCID (as in https://github.com/ropensci/codemetar/blob/master/DESCRIPTION)
 
 * web2edges/as.one.mode: check that only one direction is given; seems that some links are represented twice, even if both.directions=FALSE
 * as.one.mode/web2edges: use the number of interactions to affect the edge.weight, so that the edge.weight is the weighted mean of weights of links (or so).
@@ -65,12 +68,11 @@ This means:
 
 
 NICE TO HAVE
-* check the C4b of Gilarranz: is it the same as Opsahl's?
+* check the C4b of Gilarranz: is it the same as Opsahl's?
 * make Rcpp-version of computeModule's LPA
 * compare betweenness in sna and igraph and tnet and Pajek
 * add more networks from NCEAS and ask Jochen for other published networks; include all quantitative host-parasitoid webs (in different list as pollinators!)
 
-
 LARGELY IRRELEVANT
 large
 * sparseMatrix for Beckett algorithm?? (see email from Pierre Danelle (28 Aug 2017)