arma
,
info hereBiocCheck()
Overall summary:
Five diffusion kernels available, they can be computed from an 'igraph' object.
Diffusion implementations divided between 'diffuse_raw' for deterministic scores and 'diffuse_mc' for permutation analysis, which is parallelised. In total, seven diffusion scores are accessible through the 'diffuse' function.
Performance evaluation wrapped in the 'perf' function.
Helper functions in helpers.R (to plot diffusion scores, to check if a kernel matrix is actually a kernel, to extract largest CC from a graph)
Rcpp
latex2
to latex
graph
: must be undirectedCITATION
to the manuscriptThis version includes the requests from the reviewers
diffuse
. raw
?diffuse
?kernels
perf_wilcox
: test metrics through Wilcox testdiffuse_grid
: sweep parameters to compute diffusion scoresmetric_auc
, metric_fun
: provide basic performance metricsMetrics
to precrec
reshape2
as a dependency in Suggests
,
for its usefulness in cast
diffuse_mc
so that it accepts quantitative inputs
(but still should better be sparse)ml
and gm
now throw an error if fed with quantitative inputsThis version addresses the first reviews from Bioconductor
yeast.Rnw
to diffuStats.Rnw
gc()
browser()
Cpp
commented codediffuse.R
(gm
method)diffuStats
DESCRIPTION
fileml
could be wrongly codifiedgm
BiocCheck
not complaining anymore!diffuse
, perf
now accept kernels as inputs as well. Also, the
default Laplacian kernel is computed using kernel functionsperf
(and its core perf_eval
)
Can apply diffusion over a grid
of parameters and compute a metric using "target scores".
Unit testing written as well.pROC
dependency, added Metrics
graph_toy
to contain examples of input vector and matrix.
Updated doc and first vignette.Matrix
objects converted to base matrix in diffusion output.largest_cc
helperyeast
vignette. Takes 1:15 to build.try
to give NA
as an AUC value for bad data splits)ber_s
(both matrices are from Matrix
)gm
(missed drop = FALSE
for the single-column case)igraphdata
, the example is perfect!raw
and ber_s
will coincide if the tested nodes do not belong to
the backgrounddata-raw
raw
and ber_s
the same?
They should be similar but not the sameFirst version of the package. Includes:
igraph
backend)raw
, ml
, gm
, z
, mc
, ber_s
, ber_p
Results of devtools::check()
: 1 NOTE
(line SystemRequirements: GNU make in DESCRIPTION file)
Any scripts or data that you put into this service are public.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.