Minor issues (jasp-stats#475)

* Fix minor issues

* Remove double dependencies

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

* Improve error message

* Additional review comments

* fix unit tests

* update lockfile

* lockfile from master

---------

Co-authored-by: Copilot Autofix powered by AI <175728472+Copilot@users.noreply.github.com>

Author: JTPetter

R/doeAnalysis.R

@@ -41,10 +41,11 @@
       # avoid e-notation for small-to-moderate numbers
       mag <- floor(log10(absVal))
       sigDigits <- max(digits, mag + 1 + digits)
-      trimws(formatC(val, digits = sigDigits, format = "g", drop0trailing = TRUE))
+      formatted <- trimws(formatC(val, digits = sigDigits, format = "g", drop0trailing = TRUE))
     } else {
-      trimws(formatC(val, digits = digits, format = "g", drop0trailing = TRUE))
+      formatted <- trimws(formatC(val, digits = digits, format = "g", drop0trailing = TRUE))
     }
+    sub("^-", "\u2013", formatted)
   }, character(1))
 }
 
@@ -1527,7 +1528,7 @@ get_levels <- function(var, num_levels, dataset) {
       return()
     }
 
-    tb <- createJaspTable(gettext("ANOVA"))
+    tb <- createJaspTable(gettext("Analysis of Variance"))
     tb$addColumnInfo(name = "terms", title = gettext("Source"), type = "string")
     tb$addColumnInfo(name = "adjss", title = gettext("Sum of squares"), type = "number")
     tb$addColumnInfo(name = "df",    title = gettext("df"), type = "integer")
@@ -1967,6 +1968,16 @@ get_levels <- function(var, num_levels, dataset) {
       plotTitle <- gettextf("%1$s of %2$s vs %3$s", plotTypeString, dep, variablePairString)
       plot <- createJaspPlot(title = plotTitle, width = 560, height = 460)
       result <- jaspResults[[dep]][["doeResult"]]$object[["regression"]]
+      modelVars <- all.vars(attr(stats::terms(result[["object"]]), "variables"))
+      missingVars <- setdiff(variablePair, modelVars)
+      if (length(missingVars) > 0) {
+        plot$setError(gettextf(
+          "Could not plot because the following selected variable(s) were not included in the final model: %1$s. They might have been removed during model selection.",
+          paste(missingVars, collapse = ", ")
+        ))
+        jaspResults[[dep]][["contourSurfacePlot"]][[plotTitle]] <- plot
+        next
+      }
       if (plotType == "contourPlot") {
         plot$plotObject <- .doeContourPlotObject(result, options, dep, variablePair)
       } else if (plotType == "surfacePlot") {

R/msaGaugeRR.R

@@ -250,8 +250,8 @@ msaGaugeRR <- function(jaspResults, dataset, options, ...) {
     # Gauge r&R ANOVA Table
     if (options[["anova"]]) {
       if (is.null(jaspResults[["gaugeANOVA"]])) {
-        jaspResults[["gaugeANOVA"]] <- createJaspContainer(gettext("Gauge r&R ANOVA table"))
-        jaspResults[["gaugeANOVA"]]$dependOn(c("processVariationReference", "historicalSdValue", "report"))
+        jaspResults[["gaugeANOVA"]] <- createJaspContainer(gettext("Gauge r&R Analysis of Variance table"))
+        jaspResults[["gaugeANOVA"]]$dependOn(c("processVariationReference", "historicalSdValue", "report", "anova"))
         jaspResults[["gaugeANOVA"]]$position <- 1
       }
       jaspResults[["gaugeANOVA"]] <- .gaugeANOVA(dataset = dataset, measurements = measurements, parts = parts, operators = operators, options =  options, ready = ready, Type3 = Type3)
@@ -334,7 +334,11 @@ msaGaugeRR <- function(jaspResults, dataset, options, ...) {
     if(options[["trafficLightChart"]] & is.null(jaspResults[["trafficPlot"]] )) {
       jaspResults[["trafficPlot"]] <- createJaspContainer(gettext("Traffic light chart"))
       jaspResults[["trafficPlot"]]$position <- 9
-      jaspResults[["trafficPlot"]]$dependOn(c("trafficLightChart", "toleranceValue", "tolerance", "gaugeRRmethod", "processVariationReference", "historicalSdValue", "report"))
+      jaspResults[["trafficPlot"]]$dependOn(c("trafficLightChart", "toleranceValue", "tolerance", "gaugeRRmethod",
+                        "processVariationReference", "historicalSdValue", "studyVarianceMultiplierType",
+                        "studyVarianceMultiplierValue", "measurementLongFormat", "operatorLongFormat",
+                        "partLongFormat", "measurementsWideFormat", "operatorWideFormat",
+                        "partWideFormat", "dataFormat", "type3", "report"))
       trafficContainer <- jaspResults[["trafficPlot"]]
 
       valuesVec <- .gaugeANOVA(dataset = dataset, measurements = measurements, parts = parts, operators = operators, options =  options, ready = ready, returnTrafficValues = TRUE, Type3 = Type3)
@@ -348,11 +352,11 @@ msaGaugeRR <- function(jaspResults, dataset, options, ...) {
 
 .gaugeANOVA <- function(dataset, measurements, parts, operators, options, ready, returnPlotOnly = FALSE, returnTrafficValues = FALSE,
                         gaugeEvaluationDfOnly = FALSE, Type3 = FALSE) {
-  anovaTables <- createJaspContainer(gettext("Gauge r&R study - crossed ANOVA"))
+  anovaTables <- createJaspContainer(gettext("Gauge r&R study - crossed Analysis of Variance"))
   anovaTables$dependOn(c("anova", "gaugeRRmethod", "report"))
   anovaTables$position <- 1
 
-  anovaTable1 <- createJaspTable(title = ifelse(Type3, gettext("One-way ANOVA table"), gettext("Two-way ANOVA table with interaction")))
+  anovaTable1 <- createJaspTable(title = ifelse(Type3, gettext("One-way Analysis of Variance table"), gettext("Two-way Analysis of Variance table with interaction")))
   anovaTable1$addColumnInfo(title = gettext("Source"),       name = "source",   type = "string" )
   anovaTable1$addColumnInfo(title = gettext("df"),             name = "Df",      type = "integer")
   anovaTable1$addColumnInfo(title = gettext("Sum of squares"), name = "Sum Sq",  type = "number")
@@ -569,7 +573,7 @@ msaGaugeRR <- function(jaspResults, dataset, options, ...) {
 
     } else {
 
-      anovaTable2 <- createJaspTable(title = gettext("Two-way ANOVA table without interaction"))
+      anovaTable2 <- createJaspTable(title = gettext("Two-way Analysis of Variance table without interaction"))
       anovaTable2$addColumnInfo(title = gettext("Source"),        name = "source",   type = "string" )
       anovaTable2$addColumnInfo(title = gettext("df"),             name = "Df",      type = "integer")
       anovaTable2$addColumnInfo(title = gettext("Sum of squares"), name = "Sum Sq",  type = "number")
@@ -904,32 +908,35 @@ msaGaugeRR <- function(jaspResults, dataset, options, ...) {
 }
 
 .gaugeVarCompGraph <- function(percentContributionValues, studyVariationValues, percentToleranceValues, Type3 = FALSE) {
-  sources <- gettext(c('Gauge r&R', 'Repeat', 'Reprod', 'Part-to-part'))
+  sourceIds <- c("gaugeRR", "repeatability", "reproducibility", "partToPart")
+  sources   <- gettext(c("Gauge r&R", "Repeatability", "Reproducibility", "Part-to-part"))
   if (!all(is.na(percentToleranceValues))) {
     references <- gettextf(c('%% Contribution', '%% Study variation', '%% Tolerance'))
     values <- c(percentContributionValues, studyVariationValues, percentToleranceValues)
   } else {
     references <- gettextf(c('%% Contribution', '%% Study Variation'))
     values <- c(percentContributionValues, studyVariationValues)
   }
-  plotframe <- data.frame(source = rep(sources, length(references)),
+  plotframe <- data.frame(sourceId  = rep(sourceIds, length(references)),
+                          source    = rep(sources,   length(references)),
                           reference = rep(references, each = 4),
-                          value = values)
+                          value     = values)
   plotframe$source <- factor(plotframe$source, levels = sources)
   yBreaks <- jaspGraphs::getPrettyAxisBreaks(c(0, plotframe$value))
 
 
   if (Type3)
-    plotframe <- subset(plotframe, source != "Reprod")
+    plotframe <- subset(plotframe, sourceId != "reproducibility")
 
   p <- ggplot2::ggplot() +
     ggplot2::geom_bar(data = plotframe, mapping = ggplot2::aes(fill =  reference,  y = value, x = source),
                       position="dodge", stat = "identity") +
     jaspGraphs::themeJaspRaw() +
     jaspGraphs::geom_rangeframe() +
-    ggplot2::theme(legend.position = 'right', legend.title = ggplot2::element_blank()) +
+    ggplot2::theme(legend.position = "right", legend.title = ggplot2::element_blank(),
+             plot.margin = ggplot2::margin(5.5, 30, 5.5, 5.5, "pt")) +
     ggplot2::xlab(NULL) +
-    ggplot2::scale_y_continuous(name = "Percent", breaks = yBreaks, limits = range(c(yBreaks, plotframe$value)))
+    ggplot2::scale_y_continuous(name = gettext("Percent"), breaks = yBreaks, limits = range(c(yBreaks, plotframe$value)))
   return(p)
 }
 

R/msaGaugeRRnonrep.R

@@ -267,14 +267,19 @@ msaGaugeRRnonrep <- function(jaspResults, dataset, options, ...) {
       jaspResults[["trafficLightChart"]] <- .trafficplot(StudyVar = percentMSVar, ToleranceUsed = options[["tolerance"]],
                                                          ToleranceVar = percentMSTolerance, options = options,
                                                          ready = ready)
-      jaspResults[["trafficLightChart"]]$dependOn(c("trafficLightChart", "report"))
+      jaspResults[["trafficLightChart"]]$dependOn(c("trafficLightChart", "report", "processVariationReference",
+                        "historicalSdValue", "tolerance", "toleranceValue",
+                        "studyVarianceMultiplierType", "studyVarianceMultiplierValue",
+                        "measurementLongFormat", "operatorLongFormat", "partLongFormat",
+                        "measurementsWideFormat", "operatorWideFormat", "partWideFormat",
+                        "dataFormat"))
     }
   }
 }
 
 .gaugeRRNonRep <- function(dataset, measurements, parts, operators, options, ready, plotOnly = FALSE, trafficPlotValuesOnly = FALSE,
                            gaugeEvaluationDfOnly = FALSE) {
-  gaugeRRNonRepTables <- createJaspContainer(gettext("Gauge r&R study - nested ANOVA"))
+  gaugeRRNonRepTables <- createJaspContainer(gettext("Gauge r&R study - nested Analysis of Variance"))
   gaugeRRNonRepTables$position <- 1
 
   gaugeRRNonRepTable1 <- createJaspTable(title = gettext("Gauge r&R (nested)"))

R/msaTestRetest.R

@@ -107,7 +107,13 @@ msaTestRetest <- function(jaspResults, dataset, options, ...) {
   if(options[["trafficLightChart"]] && is.null(jaspResults[["trafficPlot"]] )) {
     jaspResults[["trafficPlot"]] <- createJaspContainer(gettext("Traffic light chart"))
     jaspResults[["trafficPlot"]]$position <- 4
-    jaspResults[["trafficPlot"]]$dependOn(c("trafficLightChart", "manualProcessSdValue", "manualProcessSd", "toleranceValue", "tolerance"))
+    jaspResults[["trafficPlot"]]$dependOn(c(
+      "trafficLightChart", "manualProcessSdValue", "manualProcessSd",
+      "toleranceValue", "tolerance",
+      "measurementLongFormat", "measurementsWideFormat",
+      "partLongFormat", "partWideFormat",
+      "operator", "dataFormat"
+    ))
     TrafficContainer <- jaspResults[["trafficPlot"]]
 
     valuesVec <- .rAndRtableRange(dataset = dataset, measurements = measurements, parts = parts, operators = operators, options =  options, jaspResults, ready = ready, GRRpercent = TRUE)

inst/help/attributesCharts.md

@@ -60,8 +60,6 @@ In the case that the data of the individual and moving Range chart (X-mR chart)
 
 - Duncan, A.J. (1986), *Quality control and industrial statistics*, Richard D. Irwin, Inc.
 - Automotive Industry Action Group, *Statistical Process Control – Reference Manual* (July 2005, 2nd Edition)
-- SKF Quality Techniques, Klerx, R., Dodson, B., and Dumont, D., QT 1 – *Process capability studies* (PUB GQ/P9 10347/1 EN – December 2021)
-- SKF Quality Techniques, Dodson, B., Lynch, D., Weidenbacher, M., and Klerx, R., QT 2 – *Statistical process control* (PUB GQS/P9 18343 EN – April 2019)
 - International Organization for Standardization, *Control charts – Part 2: Shewhart control charts*, ISO 7870-2:2023 (E)
 
 

inst/help/doeAnalysis.md

@@ -18,27 +18,27 @@ A response surface design is an experimental framework used to optimize and unde
 
 ### Assignment box
 - Response: The measured outcome variable of the process under investigation.
-- Discrete predictors: The columns corresponding to the discrete predictors in the design.
-- Continuous predictors: The columns corresponding to the continuous predictors in the design.
-- Covariates: The columns corresponding to the covariates in the design. The difference between a covariate and a continuous predictor is that the covariate will not be analysed for interaction effects and excluded from effect plots.
-- Blocks: The columns corresponding to the blocks in the design. The difference between a block variable and a discrete predictor is that the blocks will not be analysed for interaction effects and excluded from effect plots.
+- Discrete factors: The columns corresponding to the discrete factors in the design.
+- Continuous factors: The columns corresponding to the continuous factors in the design.
+- Covariates: The columns corresponding to the covariates in the design. The difference between a covariate and a continuous factor is that the covariate will not be analysed for interaction effects and excluded from effect plots.
+- Blocks: The columns corresponding to the blocks in the design. The difference between a block variable and a discrete factor is that the blocks will not be analysed for interaction effects and excluded from effect plots.
 
 ## Analysis options
 -------
 
-### Predictor levels
-For the ordering of contrasts and to identify the alpha points in a response surface design, the low and high levels of all predictors need to be specified. There are two options to handle this.
+### Factor levels
+For the ordering of contrasts and to identify the alpha points in a response surface design, the low and high levels of all factors need to be specified. There are two options to handle this.
 
-- Automatically detect low/high: Attempts to automatically detect the low and high levels by taking the minimum and maximum for continuous predictors, and by ordering discrete predictors alphabetically. For response surface designs, this might recognize alpha values as low/high levels. In this case, manual specification is needed.
+- Automatically detect low/high: Attempts to automatically detect the low and high levels by taking the minimum and maximum for continuous factors, and by ordering discrete factors alphabetically. For response surface designs, this might recognize alpha values as low/high levels. In this case, manual specification is needed.
 
-- Manually specify low/high: Shows all predictors in the analysis to allow for manual specification of the low and high levels.
+- Manually specify low/high: Shows all factors in the analysis to allow for manual specification of the low and high levels.
 
 ### Other analysis options
 
-- Use alias names: Check to assign alias names to predictors. Useful when predictors have long names.
+- Use alias names: Check to assign alias names to factors. Useful when factors have long names.
 - Show regression equation: Check to display the regression coefficients as an equation predicting the response.
-- Display result in coded units: Check to display the results in coded units. This means, all predictor levels are standardized between -1 and 1.
-- Show optimal response: Check to display the predictor levels yielding the optimal response.
+- Display result in coded units: Check to display the results in coded units. This means, all factor levels are standardized between -1 and 1.
+- Show optimal response: Check to display the factor levels yielding the optimal response.
 
 ## Model options
 -------

inst/help/doeResponseSurfaceMethodology.md

@@ -100,4 +100,4 @@ The Box-Behnken set-up consists of twelve "edge" points (shown as red dots) all
 -------
 - Dodson, L., Dodson, M., and Klerx, R. (2010), *A study of the hierarchical ordering principle and the effect heredity principle in factorial experimental designs*, 10th Annual Transaction of ENBIS (12 16 September 2010, Antwerp).
 - Box, G.E.P., and Wilson, K.B. (1951), On the experimental attainment of optimum conditions. *Journal of the Royal Statistical Society 13*(1), 1-45.
-- Dodson, B., Weidenbacher, M., Lynch D., and Klerx, R., QT 9 - Design and analysis of experiments. *SKF Quality Techniques*. (PUB GQS/P9 16083 EN - August 2015).
+- Automotive Industry Action Group (July 2005), *Statistical process control (SPC) – Reference manual*, AIAG.

inst/help/msaAttribute.md

@@ -38,7 +38,6 @@ Data can be in the form of all observations in one column ("Single column") or a
 ## References 
 -------
 - Duncan, A.J. (1986), Quality control and industrial statistics, Richard D. Irwin, Inc., and Automotive Industry Action Group (July 2005), Statistical process control (SPC) – Reference manual, AIAG.
-- Dodson, B., Lynch, D., Weidenbacher, M., & Klerx, R. (2009).*Statistical process control handbook*. SKF group. 
 ## R Packages
 -------
 - jaspGraphs

inst/help/msaGaugeLinearity.md

@@ -31,7 +31,6 @@ bias (matching between measurements and reference values) of a system.
 ## References 
 -------
 - Duncan, A.J. (1986), Quality control and industrial statistics, Richard D. Irwin, Inc., and Automotive Industry Action Group (July 2005), Statistical process control (SPC) – Reference manual, AIAG.
-- Dodson, B., Lynch, D., Weidenbacher, M., & Klerx, R. (2009).*Statistical process control handbook*. SKF group. 
 
 ## R Packages
 -------

inst/help/msaGaugeRR.md

@@ -17,13 +17,13 @@ Data can be in the form of all observations in one column ("Single column") or a
 
 ### Gauge r&R Method
 The method used in the analysis. 
-- ANOVA method: analysis is based on analysis of variance (ANOVA).
+- Analysis of Variance method: analysis is based on analysis of variance (ANOVA).
 
-### Options (ANOVA method)
+### Options (Analysis of Variance method)
 #### Analysis options 
 - Std.Deviation reference: either a historically known standard deviation (Historical process std.Deviation) or estimated from the data (Study std.Deviation).
 - Tolerance: include a value for tolerance. 
-- r&R table ANOVA method: options for the ANOVA table. 
+- r&R table Analysis of Variance method: options for the Analysis of Variance table. 
  - Alpha interaction removal: value for alpha level. 
  - Study Var. multiplier type: multiplier based on either Std.Deviation or Percent. 
  - Study Var. multiplier value: value for the multiplier. 
@@ -44,14 +44,13 @@ The method used in the analysis.
 
 ## Output 
 -------
-- ANOVA Table: Two-way anova tables with and without interactions for the input variables, repeatability, reproducibility, and total Gauge r&R.  
+- Analysis of Variance table: Two-way analysis of variance tables with and without interactions for the input variables, repeatability, reproducibility, and total Gauge r&R.  
 - Gauge r&R Variance Components: variance and contribution in percentage of the input variables, repeatability, reproducibility, and total Gauge r&R.  
 - Gauge Evaluation: the standard deviations, study variations, and percent of study variation and tolerance for input variables, repeatability, reproducibility, and total Gauge r&R.  
 
 ## References 
 -------
 - Duncan, A.J. (1986), Quality control and industrial statistics, Richard D. Irwin, Inc., and Automotive Industry Action Group (July 2005), Statistical process control (SPC) – Reference manual, AIAG.
-- Dodson, B., Lynch, D., Weidenbacher, M., & Klerx, R. (2009).*Statistical process control handbook*. SKF group. 
 
 ## R Packages
 -------