Low level fixes

DESCRIPTION

@@ -33,7 +33,8 @@ Imports:
     rjags,
     stats,
     stringr,
-    utils
+    utils,
+    rlang
 Suggests:
     knitr,
     RCurl,
@@ -46,4 +47,6 @@ LazyLoad: yes
 LazyData: true
 URL: https://github.com/nutterb/HydeNet,
 BugReports: https://github.com/nutterb/HydeNet/issues
-RoxygenNote: 7.1.0
+RoxygenNote: 7.3.2
+Encoding: UTF-8
+

NAMESPACE

@@ -53,7 +53,7 @@ export(vectorProbs)
 export(writeJagsFormula)
 export(writeNetworkModel)
 import(nnet)
-importFrom(dplyr,group_by_)
+importFrom(dplyr,group_by)
 importFrom(dplyr,summarise)
 importFrom(magrittr,"%$%")
 importFrom(magrittr,"%>%")

R/Hyde-package.R → R/HydeNet-package.R

@@ -1,21 +1,26 @@
-#' @name Hyde-package
-#' @aliases Hyde
+#' @keywords internal
+"_PACKAGE"
+
+## usethis namespace: start
 #' 
-#' @title Hydbrid Decision Networks
+#'   
+#' @aliases Hyde
 #' @description Facilities for easy implementation of hybrid Bayesian networks 
-#' using R. Bayesian networks are directed acyclic graphs representing joint 
-#' probability distributions, where each node represents a random variable and 
+#' @importFrom dplyr group_by
+#' @name HydeNet
+#' @title Hydbrid Decision Networks
+#' burdensome for inference within hybrid networks that contain a combination 
 #' each edge represents conditionality. The full joint distribution is 
-#' therefore factorized as a product of conditional densities, where each node 
+#' Econometric analyses (maximum expected utility under competing policies, 
+#' interface to constructing these networks and running inference using rjags. 
 #' is assumed to be independent of its non-descendants given information on its 
-#' parent nodes. Since exact, closed-form algorithms are computationally 
-#' burdensome for inference within hybrid networks that contain a combination 
-#' of continuous and discrete nodes, particle-based approximation techniques 
 #' like Markov Chain Monte Carlo are popular. We provide a user-friendly 
-#' interface to constructing these networks and running inference using rjags. 
-#' Econometric analyses (maximum expected utility under competing policies, 
-#' value of information) involving decision and utility nodes are also 
+#' of continuous and discrete nodes, particle-based approximation techniques 
+#' parent nodes. Since exact, closed-form algorithms are computationally 
+#' probability distributions, where each node represents a random variable and 
 #' supported.
-#'   
-
-NULL
+#' therefore factorized as a product of conditional densities, where each node 
+#' using R. Bayesian networks are directed acyclic graphs representing joint 
+#' value of information) involving decision and utility nodes are also 
+## usethis namespace: end
+NULL

R/HydePosterior.R

@@ -68,7 +68,7 @@ HydeSim <- function(cHN, variable.names, n.iter, thin=1, ...,
   if (monitor_observed)
   {
     variable.names <- 
-      if (class(cHN[["jags"]]) == "jags")
+      if (inherits(cHN[["jags"]],"jags"))
       {
         unique(c(variable.names, names(cHN$observed)))
       }
@@ -80,7 +80,7 @@ HydeSim <- function(cHN, variable.names, n.iter, thin=1, ...,
 
 
 
-  if (class(cHN$jags) == "jags")
+  if (inherits(cHN$jags,"jags"))
   {
     codas <- rjags::coda.samples(model = cHN[["jags"]], 
                           variable.names = variable.names, 

R/bindPosterior.R

@@ -50,7 +50,7 @@
 #' 
 bindSim <- function(hydeSim, relabel_factor=TRUE)
 {
-  if (class(hydeSim$codas) == "mcmc.list")
+  if (inherits(hydeSim$codas,"mcmc.list"))
   {
     bound <- 
       do.call("rbind",

R/cpt.R

@@ -87,6 +87,7 @@
 #'          reduce = FALSE)
 #' }
 #' 
+utils::globalVariables(c("sumWt"))
 cpt <- function(x, data, wt, ...)
 {
   UseMethod("cpt")
@@ -237,19 +238,19 @@ cpt_workhorse <- function(variables, dependentVar, independentVars,
 
   joint <- 
     data %>% 
-    dplyr::group_by_(.dots = ..vars) %>%  
-    dplyr::summarise_(wt = ~sum(wt))
+    group_by(!!!rlang::syms(..vars))%>%  
+    dplyr::summarise(wt = sum(wt), .groups = "drop")
 
   marginal <- 
     joint %>% 
-    dplyr::group_by_(.dots = ..independentVars) %>% 
-    dplyr::summarise_(sumWt = ~sum(wt))
+    dplyr::group_by(!!!rlang::syms(..independentVars))%>% 
+    dplyr::summarise(sumWt = sum(wt), .groups = "drop")
 
   cpt <- 
     dplyr::left_join(joint, marginal, 
                      by = independentVars) %>%
-    dplyr::mutate_(p = ~ wt / sumWt) %>% 
-    dplyr::select_(~-c(wt, sumWt)) %>%
+    dplyr::mutate(p = wt / sumWt) %>% 
+    dplyr::select(-c(wt, sumWt)) %>%
     plyr::daply(c(vars[-1], vars[1]), 
                 function(x) x$p)
 

R/print.HydeNetwork.R

@@ -86,11 +86,12 @@ print.HydeNetwork <- function(x, ...)
     {
       nodeType <- paste0(nodeType, "(", nodeParam, ")")
     }
-
-    Formula <- paste0(x[["nodeFitter"]][[node]], ": ", 
-                      deparse(x[["nodeFormula"]][[node]]) %>%
-                        trimws() %>%
-                        paste0(collapse = " "))
+    Formula <- formatNode(x[["nodeFitter"]][[node]], x[["nodeFormula"]][[node]])
+    # 
+    # Formula <- paste0(x[["nodeFitter"]][[node]], ": ", 
+    #                   deparse(x[["nodeFormula"]][[node]]) %>%
+    #                     trimws() %>%
+    #                     paste0(collapse = " "))
 
     return(paste(nodeName, nodeType, Formula, sep="\n"))
   }
@@ -111,3 +112,24 @@ print.HydeNetwork <- function(x, ...)
             }))
   cat(paste0("\n\n", nodeSummaries, "\n"))
 }
+
+#' @name formatNode
+#' 
+#' @param fitter fitter use for formula.
+#' @param formula formula use to define node in network.
+#' @title Helper function for formatting nodeFitter and nodeFormula.
+#' @description Helper function for formatting nodeFitter and nodeFormula to ensure consistent behaviour.
+
+
+formatNode <- function(fitter, formula) {
+  fitter_str <- if (is.function(fitter)) {
+    deparse(substitute(fitter))
+  } else {
+    as.character(fitter)
+  }
+  formula_str <- deparse(formula) %>%
+    trimws() %>%
+    paste0(collapse = " ")
+  paste0(fitter_str, ": ", formula_str)
+}
+

R/print.HydePosterior.R

@@ -44,7 +44,7 @@
 
 print.HydeSim <- function(x, ...){
   n_distributions <- 
-    if (class(x$codas) == "mcmc.list")
+    if (inherits(x$codas,"mcmc.list"))
     {
       1 
     }
@@ -54,7 +54,7 @@ print.HydeSim <- function(x, ...){
     }
 
   n_chains <- 
-    if (class(x[["codas"]]) == "mcmc.list")
+    if (inherits(x[["codas"]],"mcmc.list"))
     {
       length(x[["codas"]]) 
     }
@@ -64,7 +64,7 @@ print.HydeSim <- function(x, ...){
     }
 
   n_iterations <- 
-    if (class(x[["codas"]]) == "mcmc.list") 
+    if (inherits(x[["codas"]], "mcmc.list")) 
     {
       nrow(x[["codas"]][[1]]) 
     }

R/rewriteHydeFormula.R

@@ -1,5 +1,5 @@
 #' @name rewriteHydeFormula
-#' @importFrom dplyr group_by_
+#' @importFrom dplyr group_by
 #' @importFrom dplyr summarise
 #' @importFrom stats as.formula
 #' @importFrom stringr str_split_fixed