Full Changelog: v1.21.0...v1.22.6

Changes in netdiffuseR version 1.19.0 (2017-10-16)

New functions and features

• rdiffnet now allows passing scalars for threshold.dist, more over, the user
  can also ask the function to just warn when there is no diffusion instead of
  returning with error.

• plot.diffnet, plot_diffnet, plot.diffnet_mentor, and plot_diffnet2 use igraph for plotting.
  Also, users can now pass "degree" for -vertex.size- (vertex.cex not used anymore),
  allowing to "automatically" scale vertices by (in/out/.)degree. Also,
  plotting arguments like vertex.* or edge.* are standarized so these match igraph.

• plot_diffnet has a new parameter: background.

• rdiffnet_multiple, a wrapper of rdiffnet, allows performing simulations studies
  by running simulating multiple diffusion networks using rdiffnet.

• exposure has a new parameter: lags. By default lags = 0, returns a lagged
  exposure matrix.

Bug fixes

• igraph_to_diffnet was failing with the graph had no weights.

• drop_isolated was not behaving well for diffnet objects.

• vertex_covariate_dist was incorrectly specified. Only the default p=2
  were OK. Now fixed and the tests/ folder includes a tests on this.

• plot_diffnet2 was not passing color.ramp to drawColorKey. Now fixed.

• plot.diffnet_mentor had a bug. Uncesary permutation of vertices was done,
  but it actually had no visible effect. Similar problem was corrected in
  diffnet_to_igraph, and other plot methods using igraph for plotting.

Other changes

• Replacing some C functions by R functions in cases in which there
  was no decrease in performance.

• plot_diffnet function now has smaller margins, so looks more appealing.

• New examples in vignettes "netdiffuseR showcase: Medical Innovations", and
  "Simulating diffusion networks: Using the rdiffnet function".

Changes in netdiffuseR version 1.18.0 (2017-07-16)

New functions

• bootnet implements network bootstrapping based on Snijders and Borgatti (1999)

• mentor_matching implements Valente and Davis (1999) Mentor matching algorithm.
  including a plot method.

• approx_geodesic an alternative to igraph::distances and sna::geodist.
  computes geodesics up to a certain number of steps and returns a sparse
  matrix.

• matrix_compare Efficiently compares two sparse matrices looking only at the valued
  cells.

• as_dgCMatrix Coerce matrix-like objects into dgCMatrix objects (sparse matrices
  from the Matrix package).

• fitbass Fits the Bass Diffusion Model to an observed vector of cumulative
  adopters. The estimation is done via stats::nls.

• netmatch and netmatch_prepare (on development) implement matching estimators
  with network data.

Bugs Fixes

• dgr returned with error when self == TRUE

• In some calls to igraph::graph_from_adj... sorting of vertices was not preserved.

• The matrix method in egonet_attrs was passing a list of vertices instead of
  the attributes. Fixed.

• transformGraphBy was returning with error when the time periods ranged other than
  1, 2, ...

New Features

• rgraph_er is now significantly faster (orders of magnitude compare to
  previous versions). rgraph_ba is faster too.

• moran now returns the sd, expected and p-value.

• exposure now receives static graphs in alt.graph with a warning.

• rewire_graph now also uses QAP. This affects directly to structu_test.

Changes in netdiffuseR version 1.17.0 (2016-11-10)

New features and changes

• The title of the package is now Analysis of Diffusion and Contagion Processes
  on Networks.
• The function struct_test now allows other types of graphs. Before it only
  supported diffnet objects.
• The function rewire_graph gains a new argument for the algorithm "swap". Now
  to ensure aperiodicity in MCMC a chance of skiping a rewire has been included.
• The function n_rewires now has a default of 20 (before it was 100).
  This is based on Ray et al (2012) (more details in the manual).
• The function rgraph_ba gains a new argument, self=TRUE. By default behaves
  as before following Bollabas, but now can deviate to generate graphs with no
  autolinks.
• In rgraph_ba, the argument eta allows implementing De Almeida et al. (2013)
  Scale-free homophilic networks.
• The functions exposure and dgr are now pure R code (C functions were
  replace since there were no significant speed gains).
• diffnet class objects now have two new meta-values: name and behavior.
• Elements -graph-, -toa-, -adopt- and -cumadopt- in diffnet class objects
  have lost their dimnames (more efficient storage).
• classify_adopters now always includes Non-Adopters.

New functions

• vertex_covariate_dist computes distances between vertices using both the graph
  and a matrix of length nxK.
• vertex_mahalanobis_dist computes mahalanobis distance between vertices (as above).
• struct_test_asymp an asymptotic approximation of struct_test (not recomended).
• ego_variance computes a pseudo variance at the ego level (aux function for
  struct_test).
• transformGraphBy applies a function that transforms a graph considering
  structural zeros given by groups. Similar to the idea of the -by- option
  in struct_equiv.
• read_ucinet read UCINET binary files (both header and graph file). Still
  work in progress.
• plot.diffnet_degSeq method allows visualizing degree sequence as log-log
  plots (default).
• diag_expand creates a single adjacency matrix from a dynamic graph.
• summary.diffnet_adoptChange method generates a summary table of the df
  generated by select_egoalter.
• permute_graph permutes the values of an adjacency matrix (Conditional
  Uniform Graph).
• rewire_qap generates isomorphic graphs by "changing the labels".

Bug fixes

• ^.diffnet method was rasing to 1 power, e.g. diffnet^2 was actually
  diffnet^3.
• /.diffnet was not working.
• plot_diffnet was computing the coordinates of the cells wrongly. Most of the
  time causing adding figures outside of the plotting area.
• The c.diffnet_struct_test method was not updating the p.value.
• The function edgelist_to_adjmat was not processing correctly undirected
  graphs when the edgelist represented a lower triangular matrix.
• The function survey_to_diffnet had an issue processin dyn graph attrs returning
  errors. Now fixed.
• The function select_egoalter returned error when graph was an array.
• The method [[<-.diffnet failed when replacing a dynamic attribute with
  a NULL value (e.g. dn[["my_dyn_att"]] <- NULL).

Bug fixes

• Fixed bug in struct_equiv: When groupvar was a list (dynamic attr), the
  function returned error.
• Fixed bug in rewire_graph.array: Returned error when algorithm="swap"
• Fixed bug in rewire_graph: The option copy.first was not been applied correctly.
• In hist.diffnet_struct_test: ... now passed to hist.default.
• Fixed bug on egonet_attrs: The matrix method was returning with error.

New features and changes

• plot_infectsuscept includes 2D kernel smoother via MASS::kde2d.
• infection, susceptibility and threshold now report NA for non-adopters
  or excluded variables.
• egonets_attrs now has new argument: self.attrs allows including ego's attributes
  as part of the outcome so it can be used by the user.
• plot_diffnet now uses igraph::plot.igraph for plotting instead
• threshold gains a new argument: lags now users can define threshold as
  exposure lags time periods prior to the time of adoption. By default is 0
  so its exposure at the time of adoption.

New functions

• New method c.diffnet_struct_test: A wrapper of boot:::c.boot.
• diffusionMap computes the required matrix to be used with
  image-like functions mapping a vertex covariate given a graph structure.
• n_rewires: computes a suggested number of rewires per step in order to attain
  unbiased graph samples.
• diffnetLapply: Apply throught periods on diffnet objects.
• Several new methods for the class diffnet. Now users can apply str,
  dimnames (so colnames and rownames), t, &, |, dim and %*%.
• drawColorKey: Handy function to draw a color key in the current plot.
• classify_adopters: As in Valente (1995), depending on time of adoption, adopters
  are classified as early adopters, early majority, late majority, and laggards.
  The function introduces a new class and has methods for ftable and plot.
• rescale_vertex_igraph: Helper function to fix the size of vertex when calling
  plot.igraph so that the size is proportional to the x-axis.

Pre-release. Changes from CRAN to this point can be reviewed here: https://github.com/USCCANA/netdiffuseR/blob/c7c25c143a05b209bbb9b8b4c10b242179c7c79d/NEWS.md

Binary versions can be installed either within R:

> install.packages("netdiffuseR_1.16.3.29.zip", repos=NULL)

if windows, or, if OS X:

> install.packages("netdiffuseR_1.16.3.29.tgz", repos=NULL)

or using the command line, for the case of windows:

$ R CMD INSTALL netdiffuseR_1.16.3.29.zip