@@ -277,29 +277,10 @@ setMethod(f="chart_plot",

         P=output_value(dobj,'loadings')

         Ev=output_value(dobj,'eigenvalues')

-        # eigenvalues were square rooted when training PCA

-        Ev=Ev[,1]

-        Ev=Ev^2

-        

-        ## unscale the scores

-        #ev are the norms of scores

-        Ts=as.matrix(Ts) %*% diag(1/Ev) # these are normalised scores

-        

-        # scale scores and loadings by alpha

-        Ts=Ts %*% diag(Ev^(1-opt$scale_factor))

-        P=as.matrix(P) %*% diag(Ev^(opt$scale_factor))

-        

-        # additionally scale the loadings

-        sf=min(max(abs(Ts[,opt$components[1]]))/max(abs(P[,opt$components[1]])),

-            max(abs(Ts[,opt$components[2]]))/max(abs(P[,opt$components[2]])))

-        Ts=as.data.frame(Ts)

-        

-        rownames(Ts)=rownames(dobj$scores) # fix dimnames for SE object

-        colnames(Ts)=colnames(dobj$scores)

-        dobj$scores$data=as.data.frame(Ts) # nb object not returned, so only temporary scaling

+        sf = max(abs(Ts)) / max(abs(P)) * opt$scale_factor

         # plot

-        A=data.frame("x"=P[,opt$components[1]]*sf*0.8,"y"=P[,opt$components[2]]*sf*0.8)

+        A=data.frame("x"=P[,opt$components[1]]*sf,"y"=P[,opt$components[2]]*sf)

         C=pca_scores_plot(points_to_label=obj$points_to_label,xcol=obj$components[1],ycol=obj$components[2],factor_name=obj$factor_name)

         out=chart_plot(C,dobj)

@@ -472,8 +472,10 @@ setMethod(f="chart_plot",

 #' @include PCA_class.R

 #' @examples

 #' C = pca_scree_plot()

-pca_scree_plot = function(...) {

-    out=struct::new_struct('pca_scree_plot',...)

+pca_scree_plot = function(max_pc=15,...) {

+    out=struct::new_struct('pca_scree_plot',

+                           max_pc=max_pc,

+                           ...)

     return(out)

 }

@@ -481,11 +483,20 @@ pca_scree_plot = function(...) {

 .pca_scree_plot<-setClass(

     "pca_scree_plot",

     contains=c('chart'),

+    slots=c(max_pc='entity'),

     prototype = list(name='Scree plot',

         description=paste0('A plot of the percent variance and cumulative ',

             'percent variance for the components of a PCA model. '),

         type="line",

-        ontology="STATO:0000386"

+        ontology="STATO:0000386",

+        .params=c('max_pc'),

+        max_pc=entity(

+            name='Maximum number of components',

+            description='The maximum number of components to include in the plot.',

+            value=15,

+            type=c('numeric','integer'),

+            max_length=1

+        )

     )

 )

@@ -498,6 +509,10 @@ setMethod(f="chart_plot",

         ## percent variance

         scores=output_value(dobj,'scores')$data

         pvar=(colSums(scores*scores)/output_value(dobj,'ssx'))*100

+        

+        # only include up to max number of components

+        pvar=pvar[1:min(c(obj$max_pc,length(pvar)))]

+        

         A=data.frame("x"=1:length(pvar),"y"=c(pvar,cumsum(pvar)),"Variance"=as.factor(c(rep('Single component',length(pvar)),rep('Cumulative',length(pvar)))))

         labels=round(A$y,digits = 1)

         labels=format(labels,1)

@@ -422,8 +422,17 @@ setMethod(f="chart_plot",

         # 2D plot

         if (length(opt$components)==2) {

             A=data.frame("x"=P[,opt$components[1]],"y"=P[,opt$components[2]])

-            out=ggplot(data=A,aes_(x=~x,y=~y)) +

-                geom_point() +

+            out=ggplot(data=A,aes_(x=~x,y=~y))

+            if (opt$style=='points'){

+                out=out+

+                    geom_point(data=A,inherit.aes = FALSE,color='black',mapping = aes_(x=~x,y=~y))

+            }

+            if (opt$style=='arrows'){

+                out=out+

+                    geom_segment(data=A,inherit.aes = FALSE,color='black',mapping = aes_(x=~0,y=~0,xend=~x,yend=~y),arrow=arrow(length=unit(8,'points')))

+                

+            }

+            out= out + 

                 ggtitle('PCA Loadings', subtitle=NULL) +

                 xlab(paste0('Component ',opt$components[1])) +

                 ylab(paste0('Component ',opt$components[2])) +

@@ -432,7 +432,7 @@ setMethod(f="chart_plot",

             if (!is.null(obj$label_features)) {

-                if (obj$label_features=='rownames') {

+                if (obj$label_features=='row.names') {

                     vlabels=rownames(dobj$loadings)

                 } else {

                     vlabels=obj$label_features

@@ -1,7 +1,7 @@

 #' @eval get_description('pca_correlation_plot')

 #' @import struct

 #' @export pca_correlation_plot

-#' @include PCA_class.R

+#' @include scatter_chart_class.R PCA_class.R

 #' @examples

 #' C = pca_correlation_plot()

 pca_correlation_plot = function(components=c(1,2),...) {

@@ -72,9 +72,9 @@ setMethod(f="chart_plot",

 #' M = model_apply(M,D)

 #' C = pca_scores_plot(factor_name = 'Species')

 #' chart_plot(C,M[2])

-#'

 pca_scores_plot = function(

-    components=c(1,2),

+    xcol='PC1',

+    ycol='PC2',

     points_to_label='none',

     factor_name,

     ellipse='all',

@@ -83,9 +83,17 @@ pca_scores_plot = function(

     label_filter=character(0),

     label_factor='rownames',

     label_size=3.88,

+    components=NULL,

     ...) {

+    

+    if (!is.null(components)){

+        xcol=components[1]

+        ycol=components[2]

+    }

+    

     out=struct::new_struct('pca_scores_plot',

-        components=components,

+        xcol=xcol,

+        ycol=ycol,

         points_to_label=points_to_label,

         factor_name=factor_name,

         ellipse=ellipse,

@@ -94,6 +102,7 @@ pca_scores_plot = function(

         label_size=label_size,

         ellipse_type=ellipse_type,

         ellipse_confidence=ellipse_confidence,

+        components = components,

         ...)

     return(out)

 }

@@ -101,97 +110,17 @@ pca_scores_plot = function(

 .pca_scores_plot<-setClass(

     "pca_scores_plot",

-    contains='chart',

-    slots=c(

-        # INPUTS

-        components='entity',

-        points_to_label='enum',

-        factor_name='entity',

-        ellipse='enum',

-        ellipse_type='enum',

-        ellipse_confidence='entity',

-        label_filter='entity',

-        label_factor='entity',

-        label_size='entity'

-    ),

-    

+    contains='scatter_chart',

+    slots=c(components='entity'),

     prototype = list(name='PCA scores plot',

         description='Plots a 2d scatter plot of the selected components',

-        type="scatter",

-        .params=c('components','points_to_label','factor_name','ellipse',

-            'label_filter','label_factor','label_size','ellipse_type',

-            'ellipse_confidence'),

-        

-        components=entity(name='Components to plot',

-            value=c(1,2),

-            type='numeric',

-            description=paste0('The components selected for plotting.'),

+        components = entity(name='Components to plot',

+            value=NULL,

+            type=c('numeric','integer','NULL'),

+            description='The principal components used to generate the plot. If provided this parameter overrides xcol and ycol params.',

             max_length=2

         ),

-        

-        points_to_label=enum(name='Points to label',

-            value='none',

-            type='character',

-            description=c(

-                'none' = 'No samples labels are displayed.', 

-                "all" = 'The labels for all samples are displayed.', 

-                "outliers" = 'Labels for for potential outlier samples are displayed.'

-            ),

-            allowed=c('none','all','outliers')

-        ),

-        factor_name=ents$factor_name,

-        ellipse=enum(

-            name = 'Plot ellipses',

-            description=c(

-                "all" = paste0('Ellipses are plotted for all groups and all samples.'),

-                "group" = 'Ellipses are plotted for all groups.',

-                "none" = 'Ellipses are not included on the plot.',

-                "sample" = 'An ellipse is plotted for all samples (ignoring group)'),

-            allowed=c('all','group','none','sample'),

-            value='all'

-        ),

-        

-        ellipse_type=enum(

-            name='Type of ellipse',

-            description=c(

-                'norm' = paste0('Multivariate normal (p = 0.95)'),

-                't' = paste0('Multivariate t (p = 0.95)')

-                ),

-            value='norm',

-            type='character',

-            max_length = 1,

-            allowed=c('norm','t')

-        ),

-        

-        ellipse_confidence=entity(

-            name='Ellipse confidence level',

-            description='The confidence level for plotting ellipses.',

-            value=0.95,

-            type='numeric',

-            max_length = 1

-        ),

-        

-        label_filter=entity(

-            name='Label filter',

-            value=character(0),

-            type='character',

-            description=paste0(

-                'Labels are only plotted for the named groups. If ',

-                'zero-length then all groups are included.'

-            )

-        ),

-        label_factor=entity(name='Factor for labels',

-            description=paste0('The column name of sample_meta to use for ',

-                'labelling samples on the plot. "rownames" will use the row ',

-                'names from sample_meta.'),

-            type='character',

-            value='rownames',

-            max_length=1),

-        label_size=entity(name='Text size of labels',

-            description='The text size of labels. Note this is not in Font Units.',

-            type='numeric',

-            value=3.88,

-            max_length=1)

+        .params='components'

     )

 )

@@ -207,131 +136,40 @@ setMethod(f="chart_plot",

     definition=function(obj,dobj)

     {

-        if (obj$points_to_label=='outliers' & !(obj$ellipse %in% c('all','sample'))) {

-            warning('Outliers are only labelled when plotting the sample ellipse')

+        # if provided convert index to names

+        if (is.numeric(obj$xcol)) {

+            obj$xcol=colnames(dobj$scores)[obj$xcol]

         }

-        opt=param_list(obj)

+        if (is.numeric(obj$ycol)) {

+            obj$ycol=colnames(dobj$scores)[obj$ycol]

+        }

+        

+        # percent variance

         scores=output_value(dobj,'scores')$data

         pvar = (colSums(scores*scores)/output_value(dobj,'ssx'))*100 # percent variance

         pvar = round(pvar,digits = 2) # round to 2 decimal places

-        if (length(obj$factor_name)==1) {

-            shapes = 19 # filled circles for all samples

-        } else {

-            shapes = factor(dobj$scores$sample_meta[[obj$factor_name[2]]])

-        }

-        

-        if (obj$label_factor=='rownames') {

-            slabels = rownames(dobj$scores$sample_meta)

-        } else {

-            slabels = dobj$scores$sample_meta[[obj$label_factor]]

-        }

-        opt$factor_name=opt$factor_name[[1]] # only use the first factor from now on

-        

-        x=scores[,opt$components[1]]

-        y=scores[,opt$components[2]]

-        xlabel=paste("PC",opt$components[[1]],' (',sprintf("%.1f",pvar[opt$components[[1]]]),'%)',sep='')

-        ylabel=paste("PC",opt$components[[2]],' (',sprintf("%.1f",pvar[opt$components[[2]]]),'%)',sep='')

-        

-        # get the factor from meta data

-        opt$groups=dobj$scores$sample_meta[[opt$factor_name]]

-        

-        # add a space to the front of the labels to offset them from the points, because nudge_x is in data units

-        for (i in 1:length(slabels))

-        {

-            slabels[i]=paste0('  ',slabels[i], '  ')

-        }

-        

-        # filter by label_filter list if provided

-        if (length(obj$label_filter)>0) {

-            out=!(as.character(opt$groups) %in% obj$label_filter)

-            slabels[out]=''

-        }

-        

-        if (is(opt$groups,'factor') | is(opt$groups,'character')) {

-            plotClass= structToolbox:::createClassAndColors(opt$groups)

-            opt$groups=plotClass$class

-        }

-        

-        # build the plot

-        A <- data.frame (group=opt$groups,x=x, y=y,slabels=slabels)

-        

-        out = ggplot()

+        # axis labels

+        pc_idx = which(colnames(dobj$scores)==obj$xcol)

+        pc_idy = which(colnames(dobj$scores)==obj$ycol)

-        # add invisible sample points for ellipse

-        out = out+geom_point(data=A,aes_string(x='x',y='y'),alpha=0,show.legend=FALSE)

+        xlabel=paste("PC",pc_idx,' (',sprintf("%.1f",pvar[pc_idx]),'%)',sep='')

+        ylabel=paste("PC",pc_idy,' (',sprintf("%.1f",pvar[pc_idy]),'%)',sep='')

+        # copy inputs to scatter chart

+        C = scatter_chart()

+        opt=param_list(obj)

+        opt$components=NULL # exclude as not in scatter chart

+        param_list(C) = opt

-        if (length(obj$factor_name)==2) {

-            out=out+geom_point(data=A, aes_(x=~x,y=~y,colour=~group,shape=~shapes))

-        }   else {

-            out=out+geom_point(data=A, aes_(x=~x,y=~y,colour=~group))

-        }

-        

-        

-        out=out+

-            

-            geom_point(na.rm=TRUE) +

-            xlab(xlabel) +

-            ylab(ylabel) +

-            ggtitle('PCA Scores', subtitle=NULL)

-        

-        if (length(obj$factor_name)==2) {

-            out=out+labs(shape=obj$factor_name[[2]],colour=obj$factor_name[[1]])

-        } else {

-            out=out+labs(shape=obj$factor_name[[1]])

-        }

-        

-        if (obj$ellipse %in% c('all','group')) {

-            out = out +stat_ellipse(data=A, aes_(x=~x,y=~y,colour=~group),type=obj$ellipse_type,

-                level=obj$ellipse_confidence) # ellipse for individual groups

-        }

-        

-        if (is(opt$groups,'factor')) { # if a factor then plot by group using the colours from pmp package

-            out=out+scale_colour_manual(values=plotClass$manual_colors,

-                name=opt$factor_name)

-        }else {# assume continuous and use the default colour gradient

-            out=out+scale_colour_viridis_c(limits=quantile(opt$groups,

-                c(0.05,0.95),na.rm = TRUE),oob=squish,name=opt$factor_name)

-        }

-        out=out+structToolbox:::theme_Publication(base_size = 12)

-        # add ellipse for all samples (ignoring group)

-        if (obj$ellipse %in% c('all','sample')) {

-            out=out+stat_ellipse(type=obj$ellipse_type,mapping=aes(x=x,y=y),

-                colour="#C0C0C0",linetype='dashed',data=A,

-                level=obj$ellipse_confidence)

-        }

-        if (obj$ellipse %in% c('all','sample')) { # only do this if we plotted the sample ellipse

-            # identify samples outside the ellipse

-            build=ggplot_build(out)$data

-            points=build[[1]]

-            ell=build[[length(build)]]

-            # outlier for DatasetExperiment ellipse

-            points$in.ell=as.logical(sp::point.in.polygon(points$x,points$y,ell$x,ell$y))

-            

-            # label outliers if

-            if (opt$points_to_label=='outliers')

-            {

-                if (!all(points$in.ell))

-                {

-                    temp=subset(points,!points$in.ell)

-                    temp$group=opt$groups[!points$in.ell]

-                    temp$label=slabels[!points$in.ell]

-                    out=out+geom_text(data=temp,aes_(x=~x,y=~y,label=~label,colour=~group),size=obj$label_size,vjust="inward",hjust="inward",show.legend=FALSE)

-                    

-                }

-            }

-            # add a list of outliers to the plot object

-            out$outliers=trimws(slabels[!points$in.ell])

-        }

+        # plot

+        g = chart_plot(C,dobj$scores)

-        # label all points if requested

-        if (opt$points_to_label=='all') {

-            out=out+geom_text(data=A,aes_string(x='x',y='y',colour='group',label='slabels'),vjust="inward",hjust="inward",show.legend=FALSE)

-        }

+        # update axis labels

+        g = g + xlab(xlabel) +ylab(ylabel) +ggtitle('PCA scores')

-        return(out)

+        return(g)

     }

 )

@@ -462,7 +300,7 @@ setMethod(f="chart_plot",

         # plot

         A=data.frame("x"=P[,opt$components[1]]*sf*0.8,"y"=P[,opt$components[2]]*sf*0.8)

-        C=pca_scores_plot(points_to_label=obj$points_to_label,components=obj$components,factor_name=obj$factor_name)

+        C=pca_scores_plot(points_to_label=obj$points_to_label,xcol=obj$components[1],ycol=obj$components[2],factor_name=obj$factor_name)

         out=chart_plot(C,dobj)

         if (opt$style=='points')

@@ -283,7 +283,7 @@ setMethod(f="chart_plot",

         }

         if (obj$ellipse %in% c('all','group')) {

-            out = out +stat_ellipse(type=obj$ellipse_type,

+            out = out +stat_ellipse(data=A, aes_(x=~x,y=~y,colour=~group),type=obj$ellipse_type,

                 level=obj$ellipse_confidence) # ellipse for individual groups

         }

@@ -249,18 +249,26 @@ setMethod(f="chart_plot",

         }

         if (is(opt$groups,'factor') | is(opt$groups,'character')) {

-            plotClass= createClassAndColors(opt$groups)

+            plotClass= structToolbox:::createClassAndColors(opt$groups)

             opt$groups=plotClass$class

         }

         # build the plot

-        A <- data.frame (group=opt$groups,x=x, y=y)

+        A <- data.frame (group=opt$groups,x=x, y=y,slabels=slabels)

+        

+        out = ggplot()

+        

+        # add invisible sample points for ellipse

+        out = out+geom_point(data=A,aes_string(x='x',y='y'),alpha=0,show.legend=FALSE)

+        

         if (length(obj$factor_name)==2) {

-            out=ggplot (data=A, aes_(x=~x,y=~y,colour=~group,label=~slabels,shape=~shapes))

+            out=out+geom_point(data=A, aes_(x=~x,y=~y,colour=~group,shape=~shapes))

         }   else {

-            out=ggplot (data=A, aes_(x=~x,y=~y,colour=~group,label=~slabels))

+            out=out+geom_point(data=A, aes_(x=~x,y=~y,colour=~group))

         }

+        

+        

         out=out+

             geom_point(na.rm=TRUE) +

@@ -282,12 +290,11 @@ setMethod(f="chart_plot",

         if (is(opt$groups,'factor')) { # if a factor then plot by group using the colours from pmp package

             out=out+scale_colour_manual(values=plotClass$manual_colors,

                 name=opt$factor_name)

-        }

-        else {# assume continuous and use the default colour gradient

+        }else {# assume continuous and use the default colour gradient

             out=out+scale_colour_viridis_c(limits=quantile(opt$groups,

                 c(0.05,0.95),na.rm = TRUE),oob=squish,name=opt$factor_name)

         }

-        out=out+theme_Publication(base_size = 12)

+        out=out+structToolbox:::theme_Publication(base_size = 12)

         # add ellipse for all samples (ignoring group)

         if (obj$ellipse %in% c('all','sample')) {

             out=out+stat_ellipse(type=obj$ellipse_type,mapping=aes(x=x,y=y),

@@ -310,7 +317,8 @@ setMethod(f="chart_plot",

                 {

                     temp=subset(points,!points$in.ell)

                     temp$group=opt$groups[!points$in.ell]

-                    out=out+geom_text(data=temp,aes_(x=~x,y=~y,label=~label,colour=~group),size=obj$label_size,vjust="inward",hjust="inward")

+                    temp$label=slabels[!points$in.ell]

+                    out=out+geom_text(data=temp,aes_(x=~x,y=~y,label=~label,colour=~group),size=obj$label_size,vjust="inward",hjust="inward",show.legend=FALSE)

                 }

             }

@@ -319,9 +327,8 @@ setMethod(f="chart_plot",

         }

         # label all points if requested

-        if (opt$points_to_label=='all')

-        {

-            out=out+geom_text(vjust="inward",hjust="inward")

+        if (opt$points_to_label=='all') {

+            out=out+geom_text(data=A,aes_string(x='x',y='y',colour='group',label='slabels'),vjust="inward",hjust="inward",show.legend=FALSE)

         }

         return(out)

@@ -359,7 +359,7 @@ pca_biplot = function(

 .pca_biplot<-setClass(

     "pca_biplot",

-    contains=c('chart','stato'),

+    contains=c('chart'),

     slots=c(

         # INPUTS

         components='entity',

@@ -626,12 +626,12 @@ pca_scree_plot = function(...) {

 .pca_scree_plot<-setClass(

     "pca_scree_plot",

-    contains=c('chart','stato'),

+    contains=c('chart'),

     prototype = list(name='Scree plot',

         description=paste0('A plot of the percent variance and cumulative ',

             'percent variance for the components of a PCA model. '),

         type="line",

-        stato_id="STATO:0000386"

+        ontology="STATO:0000386"

     )

 )

@@ -681,7 +681,7 @@ pca_dstat_plot = function(number_components=2,alpha=0.05,...) {

 .pca_dstat_plot<-setClass(

     "pca_dstat_plot",

-    contains=c('chart','stato'),

+    contains=c('chart'),

     slots=c(number_components='entity',

         alpha='entity'),

     prototype = list(name='d-statistic plot',

@@ -690,7 +690,7 @@ pca_dstat_plot = function(number_components=2,alpha=0.05,...) {

             'considered to be outlying.'),

         type="bar",

         .params=c('number_components','alpha'),

-        stato_id='STATO:0000132',

+        ontology='STATO:0000132',

         number_components=entity(value = 2,

             name = 'Number of principal components',

             description = 'The number of principal components to use.',

@@ -78,6 +78,8 @@ pca_scores_plot = function(

     points_to_label='none',

     factor_name,

     ellipse='all',

+    ellipse_type='norm',

+    ellipse_confidence=0.95,

     label_filter=character(0),

     label_factor='rownames',

     label_size=3.88,

@@ -90,6 +92,8 @@ pca_scores_plot = function(

         label_filter=label_filter,

         label_factor=label_factor,

         label_size=label_size,

+        ellipse_type=ellipse_type,

+        ellipse_confidence=ellipse_confidence,

         ...)

     return(out)

 }

@@ -104,6 +108,8 @@ pca_scores_plot = function(

         points_to_label='enum',

         factor_name='entity',

         ellipse='enum',

+        ellipse_type='enum',

+        ellipse_confidence='entity',

         label_filter='entity',

         label_factor='entity',

         label_size='entity'

@@ -112,7 +118,9 @@ pca_scores_plot = function(

     prototype = list(name='PCA scores plot',

         description='Plots a 2d scatter plot of the selected components',

         type="scatter",

-        .params=c('components','points_to_label','factor_name','ellipse','label_filter','label_factor','label_size'),

+        .params=c('components','points_to_label','factor_name','ellipse',

+            'label_filter','label_factor','label_size','ellipse_type',

+            'ellipse_confidence'),

         components=entity(name='Components to plot',

             value=c(1,2),

@@ -135,13 +143,34 @@ pca_scores_plot = function(

         ellipse=enum(

             name = 'Plot ellipses',

             description=c(

-                "all" = paste0('Hotelling T2 95\\% ellipses are plotted for all groups and all samples.'),

-                "group" = 'Hotelling T2 95\\% ellipses are plotted for all groups.',

+                "all" = paste0('Ellipses are plotted for all groups and all samples.'),

+                "group" = 'Ellipses are plotted for all groups.',

                 "none" = 'Ellipses are not included on the plot.',

-                "sample" = 'A Hotelling T2 95\\% ellipse is plotted for all samples (ignoring group)'),

+                "sample" = 'An ellipse is plotted for all samples (ignoring group)'),

             allowed=c('all','group','none','sample'),

             value='all'

         ),

+        

+        ellipse_type=enum(

+            name='Type of ellipse',

+            description=c(

+                'norm' = paste0('Multivariate normal (p = 0.95)'),

+                't' = paste0('Multivariate t (p = 0.95)')

+                ),

+            value='norm',

+            type='character',

+            max_length = 1,

+            allowed=c('norm','t')

+        ),

+        

+        ellipse_confidence=entity(

+            name='Ellipse confidence level',

+            description='The confidence level for plotting ellipses.',

+            value=0.95,

+            type='numeric',

+            max_length = 1

+        ),

+        

         label_filter=entity(

             name='Label filter',

             value=character(0),

@@ -246,19 +275,24 @@ setMethod(f="chart_plot",

         }

         if (obj$ellipse %in% c('all','group')) {

-            out = out +stat_ellipse(type='norm') # ellipse for individual groups

+            out = out +stat_ellipse(type=obj$ellipse_type,

+                level=obj$ellipse_confidence) # ellipse for individual groups

         }

         if (is(opt$groups,'factor')) { # if a factor then plot by group using the colours from pmp package

-            out=out+scale_colour_manual(values=plotClass$manual_colors,name=opt$factor_name)

+            out=out+scale_colour_manual(values=plotClass$manual_colors,

+                name=opt$factor_name)

         }

         else {# assume continuous and use the default colour gradient

-            out=out+scale_colour_viridis_c(limits=quantile(opt$groups,c(0.05,0.95),na.rm = TRUE),oob=squish,name=opt$factor_name)

+            out=out+scale_colour_viridis_c(limits=quantile(opt$groups,

+                c(0.05,0.95),na.rm = TRUE),oob=squish,name=opt$factor_name)

         }

         out=out+theme_Publication(base_size = 12)

         # add ellipse for all samples (ignoring group)

         if (obj$ellipse %in% c('all','sample')) {

-            out=out+stat_ellipse(type='norm',mapping=aes(x=x,y=y),colour="#C0C0C0",linetype='dashed',data=A)

+            out=out+stat_ellipse(type=obj$ellipse_type,mapping=aes(x=x,y=y),

+                colour="#C0C0C0",linetype='dashed',data=A,

+                level=obj$ellipse_confidence)

         }

         if (obj$ellipse %in% c('all','sample')) { # only do this if we plotted the sample ellipse

@@ -1,11 +1,5 @@

-#' pca_correlation_plot class

-#'

-#' Plots the correlation between features and selected components.

-#'

+#' @eval get_description('pca_correlation_plot')

 #' @import struct

-#' @param components The principal components to plot (\code{numeric(2)})

-#' @param ... additional slots and values passed to struct_class

-#' @return struct object

 #' @export pca_correlation_plot

 #' @include PCA_class.R

 #' @examples

@@ -25,15 +19,18 @@ pca_correlation_plot = function(components=c(1,2),...) {

         # INPUTS

         components='entity'

     ),

-    prototype = list(name='Feature boxplot',

-        description='plots a boxplot of a chosen feature for each group of a DatasetExperiment.',

+    prototype = list(name='PCA correlation plot',

+        description=paste0(

+            'A plot of the correlation between the variables/features and ',

+            'the selected principal component scores. Features with high ',

+            'correlation are well represented by the selected component(s)'),

         type="boxlot",

         .params=c('components'),

         components=entity(name='Components to plot',

             value=c(1,2),

             type='numeric',

-            description='the components to be plotted e.g. c(1,2) plots component 1 on the x axis and component 2 on the y axis.',

+            description='The Principal Components used to generate the plot.',

             max_length=2

         )

@@ -65,25 +62,8 @@ setMethod(f="chart_plot",

 #################################################

 #################################################

-#' pca_scores_plot class

-#'

-#' 2d scatter plot of principal component scores.

-#'

+#' @eval get_description('pca_scores_plot')

 #' @import struct

-#' @param components The principal components to plot (\code{numeric(2)})

-#' @param points_to_label "none", "all", or "outliers" will be labelled on the plot.

-#' @param factor_name The sample_meta column name to use for colouring the points.

-#' You can provide up to two factors for this plot.

-#' @param ellipse "all" will plot all ellipses, "group" will only plot group ellipses,

-#' "none" will not plot any ellipses and "sample" will plot ellipse for all samples (ignoring group).

-#' @param label_filter Only include labels for samples in the group specified by label_filter.

-#' If zero length then all labels will be included.

-#' @param label_factor The sample_meta column to use for labelling the samples.

-#' If 'rownames' then the rownames will be used.

-#' @param label_size The text size of the labels.NB ggplot units, not font size units.

-#' Default 3.88.

-#' @param ... additional slots and values passed to struct_class

-#' @return struct object

 #' @export pca_scores_plot

 #' @include PCA_class.R

 #' @examples

@@ -137,42 +117,52 @@ pca_scores_plot = function(

         components=entity(name='Components to plot',

             value=c(1,2),

             type='numeric',

-            description='the components to be plotted e.g. c(1,2) plots component 1 on the x axis and component 2 on the y axis.',

+            description=paste0('The components selected for plotting.'),

             max_length=2

         ),

-        points_to_label=enum(name='points_to_label',

+        points_to_label=enum(name='Points to label',

             value='none',

             type='character',

-            description='("none"), "all", or "outliers" will be labelled on the plot.',

+            description=c(

+                'none' = 'No samples labels are displayed.', 

+                "all" = 'The labels for all samples are displayed.', 

+                "outliers" = 'Labels for for potential outlier samples are displayed.'

+            ),

             allowed=c('none','all','outliers')

         ),

-        factor_name=entity(name='Factor name',

-            value='factor',

-            type='character',

-            description='The column name of sample meta to use for plotting. A second column can be included to plot using symbols.',

-            max_length=2

-        ),

-        ellipse=enum(name = 'Plot ellipses',description=c(

-            '"all" will plot all ellipses',

-            '"group" will only plot group ellipses',

-            '"none" will not plot any ellipses',

-            '"sample" will plot ellipse for all samples (ignoring group)'),

+        factor_name=ents$factor_name,

+        ellipse=enum(

+            name = 'Plot ellipses',

+            description=c(

+                "all" = paste0('Hotelling T2 95\\% ellipses are plotted for all groups and all samples.'),

+                "group" = 'Hotelling T2 95\\% ellipses are plotted for all groups.',

+                "none" = 'Ellipses are not included on the plot.',

+                "sample" = 'A Hotelling T2 95\\% ellipse is plotted for all samples (ignoring group)'),

             allowed=c('all','group','none','sample'),

-            value='all'),

-        label_filter=entity(name='Label filter',

+            value='all'

+        ),

+        label_filter=entity(

+            name='Label filter',

             value=character(0),

             type='character',

-            description='Only include the param.group labels included in label_filter. If zero length then all labels will be included.'

+            description=paste0(

+                'Labels are only plotted for the named groups. If ',

+                'zero-length then all groups are included.'

+            )

         ),

         label_factor=entity(name='Factor for labels',

-            description='The column name of sample_meta to use as labels. "rownames" will use the row names from sample_meta.',

+            description=paste0('The column name of sample_meta to use for ',

+                'labelling samples on the plot. "rownames" will use the row ',

+                'names from sample_meta.'),

             type='character',

-            value='rownames'),

+            value='rownames',

+            max_length=1),

         label_size=entity(name='Text size of labels',

-            description='The text size of labels. Note this is not in Font Units. Default 3.88.',

+            description='The text size of labels. Note this is not in Font Units.',

             type='numeric',

-            value=3.88)

+            value=3.88,

+            max_length=1)

     )

 )

@@ -307,26 +297,13 @@ setMethod(f="chart_plot",

 #################################################################

 #################################################################

-#' pca_biplot_plot class

-#'

-#' 2d scatter plot of principal component scores overlaid with principal component loadings.

-#'

+#' @eval get_description('pca_biplot')

 #' @import struct

-#' @param components The principal components to plot (\code{numeric(2)})

-#' @param points_to_label "none", "all", or "outliers" will be labelled on the plot.

-#' @param factor_name The sample_meta column name to use for colouring the points.

-#' You can provide up to two factors for this plot.

-#' @param scale_factor Scaling factor to apply to loadings. Default = 0.95.

-#' @param style Plot style for loadings. Can be 'points' (default) or 'arrows'.

-#' @param label_features 'Include feature labels from this variable meta column. 

-#' Special keyword "rownames" will use the rownames of the variable_meta data.frame'

-#' @param ... additional slots and values passed to struct_class

-#' @return struct object

-#' @export pca_biplot_plot

+#' @export pca_biplot

 #' @include PCA_class.R

 #' @examples

-#' C = pca_biplot_plot(factor_name='Species')

-pca_biplot_plot = function(

+#' C = pca_biplot(factor_name='Species')

+pca_biplot = function(

     components=c(1,2),

     points_to_label='none',

     factor_name,

@@ -334,7 +311,7 @@ pca_biplot_plot = function(

     style='points',

     label_features=FALSE,

     ...) {

-    out=struct::new_struct('pca_biplot_plot',

+    out=struct::new_struct('pca_biplot',

         components=components,

         points_to_label=points_to_label,

         factor_name=factor_name,

@@ -346,9 +323,9 @@ pca_biplot_plot = function(

 }

-.pca_biplot_plot<-setClass(

-    "pca_biplot_plot",

-    contains='chart',

+.pca_biplot<-setClass(

+    "pca_biplot",

+    contains=c('chart','stato'),

     slots=c(

         # INPUTS

         components='entity',

@@ -358,43 +335,48 @@ pca_biplot_plot = function(

         style='enum',

         label_features='entity'

     ),

-    prototype = list(name='Feature boxplot',

-        description='plots a boxplot of a chosen feature for each group of a DatasetExperiment.',

+    prototype = list(name='PCA biplot',

+        description=paste0('A scatter plot of the selected principal ',

+            'component scores overlaid with the corresponding principal ',

+            'component loadings.'),

         type="boxlot",

         .params=c('components','points_to_label','factor_name','scale_factor','style','label_features'),

         components=entity(name='Components to plot',

             value=c(1,2),

             type='numeric',