define functions to make_: make_nodes(), make_network(),
make_dependent_events().set_ functions to set options: set_estimation_opt()
for algorithm options, and set_preprocess_opt() for preprocessing options.make_data() function to create a single data object with all
the data needed for estimation.history = c("pooling", "sequential", "consecutive") argument
to trans() and cycle() effect functions that define how the previous history of events is used to compute the effect.estimate_dynam(),
estimate_dynami() and estimate_rem().startTime is specified.lifecycle and cli to imports.snake_case to comply with the tidyverse
style guide and stocnet interface.goldfish-package.Rd.manynet.Makevars.in file that produced a note when check as CRAN.checkEvents() method addresses S3 generic/method consistency.GatherPreprocessing().estimate().estimate() for DyNAM models.estimate() when using
a parameter is fixed to a value.startTime and endTime bug on DyNAM and REM models
preprocessing.GatherPreprocessing(). Experimental functionality.aes_string() deprecation.C++ engine on DyNAM-rate..GlobalEnv and drop goldfishObjects().vcov(), now is possible to use stats::confint() to obtain
confidence interval for parameters estimates under asymptotic normality, and
also get them from a tidy() call.silent parameter replace by progress.
progress and verbose can be set with global options with the same name.logLik(), now is possible to use AIC() and BIC()
from stats package and make likelihood ratio test of nested models using
lmtest::lrtest().define functions, linkEvents() and print methods.
The reserved names and expected variables are explained.fig.retina from figures in vignettes and keeps default size.vignette("goldfishEffects") documentation for closure effects.C engine.opportunityListegoAlterInt(list(egoAttr, alterAttr))migraph packageClosure effects renaming.
clSender(net) -> commonSender(net)
clReceiver(net) -> commonReceiver(net)
mixedClSender(list(net1, net2)) -> mixedCommonSender(list(net1, net2))
mixedClReceiver(list(net1, net2)) -> mixedCommonReceiver(list(net1, net2))
Updated examine.outliers() and examine.changepoints() to plot more informative plots (with {ggplot2})
Added vignette teaching2.Rmd that also indicates these extensions
Added vignette teaching1.Rmd with the "How to start"
{broom}-related functions tidy() and glance()examine.outliers() and examine.changepoints() to plot more informative plots (with {ggplot2})
teaching2.Rmd that also indicates these extensionsteaching1.Rmd with the "How to start"New DyNAM-i model available to model face-to-face interaction data, collected through video or RFID badges.
Functions to transform interaction data into DyNAM-compatible objects
defineGroups_interaction(interactions_data, actors, seed.randomization)
Estimation functions for the joining and leaving rates of a DyNAM-i model
estimate(
dependent.events_interactions ~ 1 + intercept(network_interactions, joining = 1) +
ego(actors$age,joining = 1, subType = "centered") +
ego(actors$age,joining = -1, subType = "centered"),
model = "DyNAMi", subModel = "rate")
Estimation functions for the choice part of a DyNAM-i model
estimate(
dependent.events_interactions ~ diff(actors$age,subType="averaged_sum") ,
model = "DyNAMi", subModel = "choice",
estimationInit = list(opportunitiesList = opportunities))
New closure effects for model = "DyNAM" with subModel = "choice" and model = "REM".
Documentation is updated accordingly.
cycle(bilatnet)
clSender(bilatnet)
clReceiver(bilatnet)
mixedCycle(list(bilatnet, contignet))
mixedClSender(list(bilatnet, contignet))
mixedClReceiver(list(bilatnet, contignet))
print-method and update-method.tertius_diff() and node_trans() are changed to tertiusDiff() and
nodeTrans() complaining with naming convention.mixedTrans() effect is created. It requires the definition of two networks
to compute the number of two paths between nodes.
mixedTrans(list(bilatnet, contignet))
estimate() no longer give warnings in R 4.0.0.linkEvents() is refactored as a S3 method.print or summary methods for goldfish objects are developed.head() and tail() methods for nodes.goldfish, network.goldfish
and dependent.goldfish are available.tertius() and tertius_diff() effects are created. It requires the
definition of a network and an attribute.
tertius_diff(bilatnet, states$gdp)
tertius(bilatnet, states$gdp)
estimate() now estimation routine in C for model = "DyNAM" and
subModel = "choice_coordination".estimate() parameter engine through estimationInit argument control the estimation
routine used.
mod01 <- estimate(callsDependent ~ inertia + recip + trans,
model = "DyNAM", subModel = "choice",
estimationInit = list(engine = "gather_compute"))
estimate() now estimation routine in C except for model = "DyNAM" and
subModel = "choice_coordination".
New effects definition for model = "DyNAM" and subModel = "rate".
inertia fixed bug when weighted = TRUE.modelType argument in estimate() have been deprecated in favor of model
and submodel.estimate().defineDependentEvents(), defineGlobalAttribute(),
defineNetwork(), defineNodes() and linkEvents().