Tweaks from checks

Author: chrismainey

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: FunnelPlotR
 Type: Package
-Title: Funnel plots for comparing institutional performance
-Version: 0.2.9999
+Title: Funnel Plots for Comparing Institutional Performance
+Version: 0.3
 Authors@R: c(
         person("Chris", "Mainey", ,"chris.mainey@uhb.nhs.uk", role= c("aut", "cre"), 
           comment = c(ORCID ="0000-0002-3018-6171")
@@ -15,10 +15,10 @@ URL: https://chrismainey.github.io/FunnelPlotR, https://github.com/chrismainey/F
 BugReports: https://github.com/chrismainey/FunnelPlotR/issues
 Encoding: UTF-8
 LazyData: true
-Imports: ggrepel,
+Imports: dplyr,
+         ggrepel,
+         ggplot2,
          scales
-Depends: ggplot2,
-         dplyr
 RoxygenNote: 7.1.1
 Suggests: 
     testthat (>= 2.1.0),

NEWS.md

@@ -1,11 +1,11 @@
-# FunnelPlotR 0.2.9999
+# FunnelPlotR 0.3.0
 
-- Added methods for ratio of counts and proportions
-- Renamed several arguments
+- Added methods for ratios of counts and proportions
+- Renamed several arguments  -  please read the help file!
 - Added 'themes' as people don't like the ggplot grey as default
 - Broke process into much smaller functions for easier editing and speed
 - Tweaked vignette to make it work
-- Removed unecceary rlang dependency
+- Removed unnecessary rlang dependency
 
 # FunnelPlotR 0.2.3
 

R/funnel_plot.R

@@ -9,7 +9,7 @@
 #' @param group A vector of group names as character or factor.  Used to aggregate and group points on plots
 #' @param data_type A string identifying the type of data used for in the plot, the adjustment used and the reference point. One of: "SR" forindirectly standardised ratios, such SHMI, "PR" for proportions, or "RC" for ratios of counts. Default is "SR".
 #' @param title Plot title
-#' @param label_outliers Add group labels to outliers on plot. Accepted values are: 95 or 99 corresponding to 95\% or 99.8\% quantiles of the distribution. Default=99
+#' @param label_outliers Add group labels to outliers on plot. Accepted values are: 95 or 99 corresponding to 95\% or 99.8\% quantiles of the distribution. Default=99,and applies to OD limits if both OD and Poisson are used.
 #' @param Poisson_limits Draw exact Poisson limits, without overdispersion adjustment. (default=FALSE)
 #' @param OD_adjust Draw overdispersed limits using hierarchical model, assuming at group level, as described in Spiegelhalter (2012) <doi:https://doi.org/10.1111/j.1467-985X.2011.01010.x>.
 #' It calculates a second variance component ' for the 'between' standard deviation (Tau2), that is added to the 'within' standard deviation (sigma) (default=TRUE)
@@ -18,7 +18,7 @@
 #' SHMI, instead, uses log-transformation and doesn't Winsorise, but truncates the distribution before assessing overdisperison.
 #' Both methods then calculate a dispersion ratio (phi) on this altered dataset.  This ratio is then used to scale the full dataset,
 #' and the plot is drawn for the full dataset.
-#' @param winsorise_by Proportion of the distribution for winsorisation/truncation. Default is 10 \% (0.1).  Note, this is applied in a two-sided
+#' @param trim_by Proportion of the distribution for winsorisation/truncation. Default is 10 \% (0.1).  Note, this is applied in a two-sided
 #' fashion, e.g. 10\% refers to 10\% at each end of the distribution (20\% winsorised/truncated)
 #' @param multiplier Scale relative risk and funnel by this factor. Default to 1, but 100 sometime used, e.g. in some hospital mortality ratios.
 #' @param x_label Title for the funnel plot x-axis.  Usually expected deaths, readmissions, incidents etc.
@@ -71,7 +71,7 @@
 
 
 funnel_plot <- function(numerator, denominator, group, data_type = "SR", label_outliers = 99,
-                            Poisson_limits = FALSE, OD_adjust = TRUE, sr_method = "SHMI", winsorise_by = 0.1,
+                            Poisson_limits = FALSE, OD_adjust = TRUE, sr_method = "SHMI", trim_by = 0.1,
                             title="Untitled Funnel Plot", multiplier = 1, x_label = "Expected",
                             y_label ,xrange = "auto", yrange = "auto",
                             return_elements=c("plot", "data", "limits"), theme = funnel_clean()){
@@ -131,7 +131,11 @@ funnel_plot <- function(numerator, denominator, group, data_type = "SR", label_o
   mod_plot_agg <- transformed_zscore(mod_plot_agg=mod_plot_agg, data_type = data_type, sr_method = sr_method)
 
   # Winsorise or truncate depending on method
-  mod_plot_agg <- winsorisation(mod_plot_agg = mod_plot_agg, data_type=data_type, sr_method = sr_method, winsorise_by=winsorise_by)
+  if(data_type=="SR" & sr_method=="SHMI"){
+    mod_plot_agg <- truncation(mod_plot_agg = mod_plot_agg, trim_by=trim_by)
+  } else {
+    mod_plot_agg <- winsorisation(mod_plot_agg = mod_plot_agg, trim_by=trim_by)
+  }
 
   n <- as.numeric(sum(!is.na(mod_plot_agg$Wuzscore)))
   # Calculate Phi (the overdispersion factor)

R/od_adjustment.R

@@ -66,44 +66,63 @@ transformed_zscore<-function(mod_plot_agg=mod_plot_agg, data_type = "SR", sr_met
 
 #' Winsorisation function
 #'
-#' @description Internal function to perform the winsorisation or truncation.
+#' @description Internal function to perform the Winsorisation.
 #'
 #' @param mod_plot_agg Aggregated model input data
-#' @param data_type Type of data for adjustment and plotting: Indirectly Standardised ratio (\"SR\"), proportion (\"PR\"), or ratio of counts (\"RC\").
-#' @param sr_method Adjustment method for standardised ratios, can take the value \"SHMI\" or \"CQC\". \"SHMI\" is default. Not relevant to PR or RC data types
-#' @param winsorise_by The amount to winsorise\/truncate the distribution by, prior to transformation. 0.1 means 10\% (at each end).
+#' @param trim_by The amount to Winsorise the distribution by, prior to transformation. 0.1 means 10\% (at each end).
 #'
-#' @return A data.frame with winsorized z-scores returned added
+#' @return A data.frame with winsorised z-scores returned added
 #' @keywords internal
 #' 
 #' 
-winsorisation <- function(mod_plot_agg = mod_plot_agg, data_type=data_type, sr_method = "SHMI", winsorise_by = 0.1){
+winsorisation <- function(mod_plot_agg = mod_plot_agg, trim_by = 0.1){
 
-  lz <- quantile(x = mod_plot_agg$Uzscore, winsorise_by, na.rm = TRUE)
-  uz <- quantile(x = mod_plot_agg$Uzscore, (1 - winsorise_by), na.rm = TRUE)
+  lz <- quantile(x = mod_plot_agg$Uzscore, trim_by, na.rm = TRUE)
+  uz <- quantile(x = mod_plot_agg$Uzscore, (1 - trim_by), na.rm = TRUE)
   mod_plot_agg$winsorised = ifelse(mod_plot_agg$Uzscore > lz & mod_plot_agg$Uzscore < uz, 0, 1)
 
-  if(data_type == "SR" & sr_method == "SHMI"){
-    
-    #mod_plot_agg$Wuzscore[mod_plot_agg$winsorised == 1] <- mod_plot_agg$Uzscore
-    mod_plot_agg$Wuzscore <-ifelse(mod_plot_agg$winsorised == 0, mod_plot_agg$Uzscore, NA)
-  
-  } else {
-  
-  #if(sr_method == "CQC"){
-
-    mod_plot_agg$Wuzscore = ifelse(mod_plot_agg$Uzscore < lz
+  mod_plot_agg$Wuzscore = ifelse(mod_plot_agg$Uzscore < lz
                                    , lz
                                    , ifelse(mod_plot_agg$Uzscore > uz
                                           , uz
                                           , mod_plot_agg$Uzscore))
-  }
+  
+  return(mod_plot_agg)
+}
+
+#' Truncation function for NHSD method
+#'
+#' @description Internal function to perform the truncation.
+#'
+#' @param mod_plot_agg Aggregated model input data
+#' @param trim_by The amount to truncate the distribution by, prior to transformation. 0.1 means 10\% (at each end).
+#'
+#' @return A data.frame with truncated z-scores added
+#' @keywords internal
+#' 
+#' 
+truncation <- function(mod_plot_agg = mod_plot_agg, trim_by = 0.1){
+  
+  # How many groups for truncation
+  k <- 1/trim_by
+  maxk<- k-1
+  mink<-min(k/k)-1
+  
+  mod_plot_agg$rk <- rank(mod_plot_agg$Uzscore, ties.method = "average")
+  mod_plot_agg$sp <- floor(mod_plot_agg$rk  * k / (length(mod_plot_agg$rk) + 1))
+  
+  mod_plot_agg$truncated = ifelse(mod_plot_agg$sp > mink & mod_plot_agg$Uzscore < maxk, 0, 1)
+  
+  mod_plot_agg$Wuzscore = ifelse(mod_plot_agg$truncated == 1 
+                                , NA
+                                , mod_plot_agg$Uzscore)
+  
   return(mod_plot_agg)
 }
 
 
 
-#' Transformation function for z-scoring
+#' Calculate overdispersion ratio
 #'
 #' @description Internal function to perform the transformations for data types.
 #'

README.Rmd

@@ -21,7 +21,7 @@ knitr::opts_chunk$set(
 
 
 
-# Funnel plots for comparing institutional performance <img src="man/figures/logo.png" width="160px" align="right" />
+# Funnel Plots for Comparing Institutional Performance <img src="man/figures/logo.png" width="160px" align="right" />
 
 <!-- badges: start -->
 [![Travis build status](https://travis-ci.org/chrismainey/FunnelPlotR.svg?branch=master)](https://travis-ci.org/chrismainey/FunnelPlotR)
@@ -49,7 +49,7 @@ Methods are based on those presented in Spiegelhalter's papers and the Care Qual
 This uses a log-transformation and truncation of the distribution for calculating overdispersion, whereas Spiegelhalter's methods use a square-root and Winsorisation.
 
 
-This package was originally developed for use in the author's PhD project, but published on github in case it's of use for others.
+This package was originally developed for use in the author's PhD project, but published on Github in case it's of use for others.
 
 Contributions are welcome.  Please note that the 'FunnelPlotR' project is released with a
 [Contributor Code of Conduct](https://chrismainey.github.io/FunnelPlotR/CODE_OF_CONDUCT.html).

README.md

@@ -1,5 +1,5 @@
 
-# Funnel plots for comparing institutional performance <img src="man/figures/logo.png" width="160px" align="right" />
+# Funnel Plots for Comparing Institutional Performance <img src="man/figures/logo.png" width="160px" align="right" />
 
 <!-- badges: start -->
 
@@ -55,7 +55,7 @@ calculating overdispersion, whereas Spiegelhalter’s methods use a
 square-root and Winsorisation.
 
 This package was originally developed for use in the author’s PhD
-project, but published on github in case it’s of use for others.
+project, but published on Github in case it’s of use for others.
 
 Contributions are welcome. Please note that the ‘FunnelPlotR’ project is
 released with a [Contributor Code of

cran-comments.md

@@ -1,8 +1,7 @@
 ## Release summary
-This is a minor bug-fix for the FunnelPlotR where some plot limits were incorrectly coloured3.
+This is a major release for the FunnelPlotR package, adding more methods, changes to uer interface, and removing dependencies.
 
 ## Test environments
-* local windows 7, R 3.6.1
 * local windows 10, R 3.6.1
 * Windows dev on Win-builder
 * ubuntu 16.04.6 LTS (on travis-ci), R 3.6.1

dev/tests.R

@@ -3,12 +3,12 @@ library(COUNT)
 library(ggplot2)
 
 # Setup
-data_type = "PR"
+data_type = "SR"
 label_outliers = 95
 Poisson_limits = FALSE
 OD_adjust = TRUE
 sr_method = "SHMI"
-Winsorise_by = 0.1
+trim_by = 0.1
 title="Untitled Funnel Plot"
 multiplier = 1
 x_label = "Expected"
@@ -31,7 +31,7 @@ medpar$prds<- predict(mod, type="response")
 
 # Draw plot, returning just the plot object
 fp2<-funnel_plot(denominator=medpar$prds,numerator=medpar$los, 
-                 group = medpar$provnum, label_outliers = 99,
+                 group = medpar$provnum, label_outliers = 95,
                  Poisson_limits = TRUE)
 
 fp2