0.6.2 - fix DataSummary issues

DESCRIPTION

@@ -1,8 +1,8 @@
 Package: DataMan
 Type: Package
 Title: Data Summary, One-way Interactive Plot and Model Visualisation Analysis
-Version: 0.6.1
-Date: 2018-6-17
+Version: 0.6.2
+Date: 2018-7-18
 Author: Sixiang Hu <[email protected]>
 Maintainer: Sixiang Hu <[email protected]>
 Description: DataMan package provides functions for data summary, one-way plot, and model assess functionalities with 2D and 3D interactive visualisation.

NAMESPACE

@@ -35,6 +35,7 @@ export(resiPlot)
 export(rocPlot)
 export(tree2data)
 export(varComp)
+import(data.table)
 importFrom(ROCR,performance)
 importFrom(ROCR,prediction)
 importFrom(Rcpp,evalCpp)
@@ -45,8 +46,6 @@ importFrom(checkmate,testNumeric)
 importFrom(data.table,":=")
 importFrom(data.table,as.data.table)
 importFrom(data.table,data.table)
-importFrom(data.table,dcast)
-importFrom(data.table,key)
 importFrom(data.table,melt)
 importFrom(data.table,set)
 importFrom(data.table,setDT)
@@ -69,5 +68,6 @@ importFrom(plotly,subplot)
 importFrom(randomForest,getTree)
 importFrom(stats,chisq.test)
 importFrom(stats,na.omit)
+importFrom(stats,var)
 importFrom(stats,weighted.mean)
 useDynLib(DataMan)

NEWS.md

@@ -1,3 +1,9 @@
+DataMan 0.6.2
+-------------------------------------------------------------
+BUG FIXES AND MINOR IMPROVEMENTS
+* `DataSummary` Fix issue #48: unique value count is not correct when `NA` included.
+* `DataSummary` Fix issue #49: mean value is not correct when `NA` included.
+
 DataMan 0.6.1
 -------------------------------------------------------------
 NEW FEATURES

R/DataSummary.R

@@ -1,7 +1,7 @@
 #' DataSummary
 #'
 #' @description function gives summary of the dataset.
-#' @usage DataSummary(data,missing=list(NA),wt=NULL,entropy=FALSE,
+#' @usage DataSummary(data,missing=list(NA,NaN,Inf,-Inf),wt=NULL,entropy=FALSE,
 #' entropy_method="emp",sparkline=FALSE)
 #' @param data This could be data frame, a vector, matrix or data.table.
 #' @param missing list of possible missing values.
@@ -22,15 +22,15 @@
 #' @importFrom checkmate testCharacter testNumeric
 #' @export DataSummary
 #' @examples
-#' DataSummary(mtcars,missing=list(NA,".","Unknown",-1))
+#' DataSummary(mtcars,missing=list(NA,NaN,Inf,-Inf,".","Unknown",-1))
 
-DataSummary <- function(data,missing=list(NA,NaN,Inf),wt=NULL,entropy=FALSE,entropy_method="emp",sparkline=FALSE){
+DataSummary <- function(data,missing=list(NA,NaN,Inf,-Inf),wt=NULL,entropy=FALSE,entropy_method="emp",sparkline=FALSE){
   UseMethod("DataSummary",data)
 }
 
 #' @export
 #' @rdname DataSummary
-DataSummary.data.frame <- function(data,missing=list(NA),wt=NULL,
+DataSummary.data.frame <- function(data,missing=list(NA,NaN,Inf,-Inf),wt=NULL,
                                    entropy=FALSE,entropy_method="emp",sparkline=FALSE){
   setDT(data)
   DataSummary.data.table(data,missing,wt,entropy,entropy_method,sparkline)
@@ -39,7 +39,7 @@ DataSummary.data.frame <- function(data,missing=list(NA),wt=NULL,
 #' @export
 #' @rdname DataSummary
 #' 
-DataSummary.data.table <- function(data,missing=list(NA),wt=NULL,
+DataSummary.data.table <- function(data,missing=list(NA,NaN,Inf,-Inf),wt=NULL,
                                    entropy=FALSE,entropy_method="emp",sparkline=FALSE){
 
   if(is.null(wt)) weight <- rep(1,nrow(data))

R/interPlot.R

@@ -14,7 +14,7 @@
 #' This is really useful when modeller wants to assess the interaction terms.
 #' Hence the `z` can be actual response data or model predictions.
 #' @author Sixiang Hu
-#' @importFrom data.table data.table setkey key := dcast
+#' @import data.table
 #' @importFrom plotly plot_ly layout add_surface %>%
 #' @export interPlot
 #' @examples
@@ -72,4 +72,4 @@ interPlot <- function(x,y,z,weight=NULL,exposure=NULL,numGrpx=10,numGrpy=10,
 }
 
 #global variable
-globalVariables(c("e","w"))
+globalVariables(c("e","w","."))

R/resiPlot.r

@@ -21,8 +21,9 @@
 #' heterogeneity
 #' 
 #' @author Sixiang Hu
-#' @importFrom data.table setkey data.table
+#' @import data.table
 #' @importFrom plotly plot_ly add_histogram2dcontour add_trace %>% subplot
+#' @importFrom stats var
 #' @export resiPlot
 #' @examples
 #' set.seed(1L)
@@ -88,4 +89,4 @@ resiPlot <- function(act,pred,weight=NULL,exposure=NULL,bucket=20){
 }
 
 #global variable
-globalVariables(c("x","e","w"))
+globalVariables(c("x","e","w","."))