Add interrater reliability and update 1.0.0

.gitignore

@@ -0,0 +1,4 @@
+.Rproj.user
+.Rhistory
+.RData
+.Ruserdata

DESCRIPTION

@@ -1,25 +1,23 @@
 Package: seolmatrix
 Type: Package
 Title: Correlations suite for jamovi
-Version: 0.0.3
+Version: 1.0.0
 Date: 2020-06-17
 Author: Hyunsoo Seol
 Maintainer: Hyunsoo Seol <[email protected]>
-Description:This module is a tool for calculating correlations, such as Pearson, Partial, Point-Biserial,Tetrachoric and allows users to produce 
+Description: This module is a tool for calculating correlations, such as Pearson, Partial, Point-Biserial,Tetrachoric, rater, and allows users to produce 
   Gaussian Graphical Model.
 License: GPL (>= 2)
 Encoding: UTF-8
 LazyData: true
-RoxygenNote: 7.1.0
+RoxygenNote: 7.1.1
 Imports: 
     janitor,
     jmvcore (>= 1.0.8),
     knitr,
     R6,
-    magrittr,
-    tidyverse,
     stats,
     qgraph,
-    psych
-
-
+    psych,
+    lpSolve,
+    irr

NAMESPACE

@@ -4,8 +4,15 @@ export(dichotomous)
 export(dichotomousClass)
 export(partial)
 export(partialClass)
+export(rater)
+export(raterClass)
+import(irr)
 import(jmvcore)
+import(lpSolve)
 import(psych)
 import(qgraph)
 importFrom(R6,R6Class)
+importFrom(irr,kappam.light)
 importFrom(magrittr,"%>%")
+importFrom(psych,partial.r)
+importFrom(psych,polychoric)

R/00jmv.R

@@ -22,4 +22,10 @@
         `year`=2012, 
         `title`="qgraph: Network Visualizations of Relationships in Psychometric Data", 
         `publisher`="[R package]. Retrieved from https://CRAN.R-project.org/package=qgraph", 
-        `url`="https://CRAN.R-project.org/package=qgraph"))
+        `url`="https://CRAN.R-project.org/package=qgraph"), 
+    `irr`=list(
+        `type`="software", 
+        `author`="Matthias Gamer and Jim Lemon and Ian Fellows Puspendra Singh", 
+        `year`=2019, 
+        `title`="irr: Various Coefficients of Interrater Reliability and Agreement", 
+        `url`="https://CRAN.R-project.org/package=irr"))

R/dichotomous.b.R

@@ -40,7 +40,9 @@ self$results$instructions$setContent(
             <p>- The input dataset require dichotomous data with the type of <b>numeric-continuous</b> in jamovi.</p>
             <p>- Just highlight the variables and click the arrow to move it across into the 'Variables' box.</p>
             
-            <p>If you encounter any errors, or have questions, please e-mail me: [email protected]</a></p>
+             <p>- Feature requests and bug reports can be made on my <a href='https://github.com/hyunsooseol/seolmatrix/'  target = '_blank'>GitHub</a></p>
+
+            <p>If you have any questions, please e-mail me: [email protected]</a></p>
             </div>
             </body>
             </html>"

R/dichotomous.h.R

@@ -55,7 +55,8 @@ dichotomousResults <- if (requireNamespace('jmvcore')) R6::R6Class(
                 options=options,
                 name="instructions",
                 title="Instructions",
-                visible=TRUE))
+                visible=TRUE,
+                refs="seolmatrix"))
             self$add(jmvcore::Table$new(
                 options=options,
                 name="matrix",
@@ -77,7 +78,7 @@ dichotomousResults <- if (requireNamespace('jmvcore')) R6::R6Class(
             self$add(jmvcore::Image$new(
                 options=options,
                 name="plot",
-                title="Show Gaussian Graphical Model",
+                title="Gaussian Graphical Model",
                 width=500,
                 height=500,
                 renderFun=".plot",

R/partial.b.R

@@ -49,8 +49,9 @@ partialClass <- if (requireNamespace('jmvcore'))
 
             <p> Note: When One variable is dichotomous, the other is continuous, the result table is equivalent to a <b>Point-biserial correlation</b>.</P>
 
-            <p>If you encounter any errors, or have questions, please e-mail me: [email protected]</a></p>
+            <p>- Feature requests and bug reports can be made on my <a href='https://github.com/hyunsooseol/seolmatrix/'  target = '_blank'>GitHub</a></p>
 
+            <p>If you have any questions, please e-mail me: [email protected]</a></p>
             </div>
             </body>
             </html>")

R/rater.b.R

@@ -0,0 +1,224 @@
+
+# This file is a generated template, your changes will not be overwritten
+
+#' Interrater Analysis
+#' @importFrom R6 R6Class
+#' @import jmvcore
+#' @importFrom irr kappam.light
+#' @import irr
+#' @import lpSolve
+#' @import qgraph
+#' @import psych
+#' @importFrom psych polychoric
+#' @importFrom  psych partial.r
+#' @export
+
+
+raterClass <- if (requireNamespace('jmvcore')) R6::R6Class(
+    "raterClass",
+    inherit = raterBase,
+    private = list(
+
+#==========================================================
+        .init = function() {
+
+
+            if(is.null(self$data) | is.null(self$options$vars)){
+                self$results$instructions$setVisible(visible = TRUE)
+
+            }
+
+            self$results$instructions$setContent(
+                "<html>
+            <head>
+            </head>
+            <body>
+            <div class='instructions'>
+            <p>Welcome to Interrater reliability.</p>
+            
+            <p><b>To get started:</b></p>
+            
+            <p>- Just highlight the variables(raters) and click the arrow to move it across into the 'Variables' box.</p>
+            <p>- To produce Gaussian Graphical Model, partial correlations are calculated from a input matrix of polychoric correlations.</p> 
+            
+            <p>- Feature requests and bug reports can be made on my <a href='https://github.com/hyunsooseol/seolmatrix/'  target = '_blank'>GitHub</a></p>
+
+            <p>If you have any questions, please e-mail me: [email protected]</a></p>
+            </div>
+            </body>
+            </html>"
+            )
+
+        },
+
+        #======================================++++++++++++++++++++++
+        .run = function() {
+
+            # get variables-------
+
+            data <- self$data
+
+            vars <- self$options$get('vars')
+
+
+            # Ready--------
+
+            ready <- TRUE
+
+            if (is.null(self$options$vars) ||
+                length(self$options$vars) < 2)
+
+                ready <- FALSE
+
+            if (ready) {
+
+                data <- private$.cleanData()
+
+                results <- private$.compute(data)
+
+                # populate rater table-----
+
+                private$.populateRaterTable(results)
+
+
+                # prepare plot-----
+
+                  private$.prepareGgmPlot(data)
+
+            }
+        },
+
+  # compute results=====================================================
+
+        .compute = function(data) {
+
+            # get variables------
+
+            data <- self$data
+
+            vars <- self$options$get('vars')
+
+
+             # for(v in vars)
+             #   data[[v]] <- jmvcore::toNumeric(data[[v]])
+
+            # compute Light's Kappa-----
+
+            res<- irr::kappam.light(ratings=data)
+
+            # get subjects-------
+
+            n <- res$subjects
+
+            # get raters--------
+
+            rater <- res$raters
+
+            # get statistic------------
+
+            statistic <- res$value
+
+            # z value----------------
+
+            z <- res$statistic
+
+            # p value-------------------
+
+            p <- res$p.value
+
+            results <-
+                list(
+                    'n' = n,
+                    'rater' = rater,
+                    'statistic' = statistic,
+                    'z' = z,
+                    'p' = p
+
+                )
+
+        },
+
+
+ # populate rater table-----
+
+ .populateRaterTable = function(results) { 
+
+          table <- self$results$interrater
+
+            n <- results$n
+            rater <- results$rater
+            statistic <- results$statistic
+            z <- results$z
+            p <- results$p
+
+
+            row <- list()
+
+            row[['n']] <- n
+            row[['rater']] <- rater
+            row[['statistic']] <- statistic
+            row[['z']] <- z
+            row[['p']] <- p
+
+            table$setRow(rowNo = 1, values = row)
+
+ },
+
+
+#### Plot functions ----
+
+.prepareGgmPlot = function(data) {
+
+  # Gaussian Graphical Models with polychoric correlation----------------
+
+  poly <- psych::polychoric(data)$rho
+  partial <- psych::partial.r(poly)
+
+  # Prepare Data For Plot -------
+  image <- self$results$plot
+  image$setState(partial)
+
+
+
+},
+
+
+#================================================================
+
+.plot = function(image, ...) {
+  ggm <- self$options$ggm
+
+  if (!ggm)
+    return()
+
+
+  partial <- image$state
+
+  plot <- qgraph(partial, layout = "spring", details = TRUE)
+
+  print(plot)
+  TRUE
+},
+
+
+        #### Helper functions =================================
+
+        .cleanData = function() {
+            items <- self$options$vars
+
+            data <- list()
+
+            for (item in items)
+                data[[item]] <- jmvcore::toNumeric(self$data[[item]])
+
+            attr(data, 'row.names') <- seq_len(length(data[[1]]))
+            attr(data, 'class') <- 'data.frame'
+            data <- jmvcore::naOmit(data)
+
+            return(data)
+        }
+    )
+)
+
+
+