classdef (CaseInsensitiveProperties = true) boxPlot < iosr.statistics.statsPlot %BOXPLOT Draw a box plot % % Use this class to draw box plots. The class provides a number of % options not included in Matlab's boxplot function. % % For each box, the central mark is the median, the box extends % vertically between the 25th and 75th percentiles, the whiskers extend % to the most extreme data that are not considered outliers, and the % outliers are plotted individually. % % BOXPLOT can draw boxes for data in Y for an arbitrary number of % dimensions. If Y is an N-by-P-by-G-by-I-by-J... array then G*I*J*... % boxes are plotted hierarchically for each column P; i.e. J boxes are % plotted for each index of I, and J*I boxes are plotted for each index % of G. % % IOSR.STATISTICS.BOXPLOT properties: % Plotting options: % addPrctiles - Show additional percentiles using markers and % labels. The property should be a vector of % percentiles; each percentile will be plotted % for each box. The property is empty by % default. % addPrctilesColor - Specify the marker color for the additional % percentile markers. The property should be a % cell array of strings indicating the color % for each percentile; the colors will be % repeated for each box. The default is black. % addPrctilesLabels - Specify labels for the additional % percentiles. The property should be a cell % array of strings. By defualt no label % is shown. % addPrctilesMarkers - Specify markers for the additional % percentiles. The property should be a cell % array of strings indicating the shape of each % percentile; the markers will be repeated for % each box. The default is '*'. % addPrctilesSize - Specify the marker size for the additional % percentile markers. The property should be a % numeric vector indicating the size for each % percentile; the markers will be repeated for % each box. The default is 6. % addPrctilesTxtSize - Specify the font size of the additional % percentile labels. The property should be a % numeric scalar. The default is 9. % boxAlpha - The transparency of the boxes (0 is % transparent, 1 is opaque). The default is 1. % boxColor - Fill color of the box. The setting can be a % color specifier (3-element vector or single % character), 'auto', 'none', or a handle to a % colormap function (e.g. @gray); 'auto' means % that Matlab's default colors are used. The % default is 'none'. % boxWidth - The width of the box. The setting can be % 'auto' (meaning that the width is determined % automatically) or a scalar (specifying the % width in x-axis units). The default is % 'auto'. % groupLabelFontSize - Specify the font size of the group labels % when the 'style' option is set to % 'hierarchy'. The default is 9. % groupLabelHeight - Specify the height of the area reserved for % group labels when the 'style' option is set % to 'hierarchy'. The height is determined % automatically ('auto') or can be specified in % scale units. The default is 'auto'. % groupLabels - If the 'style' option is set to 'hierarchy' % then these labels will be used to label the % boxes for each x-group. The parameter should % be specified as a cell vector whereby the Nth % element contains a vector of length % SIZE(Y,N+2). By default, the labels are % empty. % groupWidth - Specify the proportion of the x-axis interval % across which each x-group of boxes should be % spread. The default is 0.75. % limit - Mode indicating the limits that define % outliers. When set to '1.5IQR' or '3IQR', the % min and max values are Q1-Z*IQR and Q3+Z*IQR, % where Z = 1.5 or 3 respectively. When set to % 'none', the min and max values are the min % and max of the data (in this case there will % be no outliers). The property may also be % specified as a two-element vector determining % the limits as percentiles (e.g. [5,95]). The % default is '1.5IQR'. % lineColor - Color of the box outline and whiskers. The % default is 'k'. % lineStyle - Style of the whisker line. The default is % '-'. % lineWidth - Width, in points, of the box outline, whisker % lines, notch line, violin, and outlier marker % edges. The default is 1. % meanColor - Color of the mean marker when showMean=true. % The default is 'auto' (see boxColor). % meanMarker - Marker used for the mean when showMean=true. % The default is '+'. % meanSize - Size, in point units, of the mean markers % when showMean=true. The default is 6. % medianColor - Color of the median line. The default is % 'auto' (see boxColor). % method - The method used to calculate the quantiles, % labelled according to % http://en.wikipedia.org/wiki/Quantile. The % default is 'R-8' whereas the default for % Matlab is 'R-5'. See % IOSR.STATISTICS.QUANTILE. % notch - Logical value indicating whether the box % should have a notch. The notch is centred on % the median and extends to �1.58*IQR/sqrt(N), % where N is the sample size (number of non-NaN % rows in Y). Generally if the notches of two % boxes do not overlap, this is evidence of a % statistically significant difference between % the medians. The default is false. % notchDepth - Depth of the notch as a proportion of half % the box width. The default is 0.4. % notchLine - Logical value specifying whether to draw a % horizontal line in the box at the extremes of % the notch. (May be specified indpendently of % 'notch'.) The default is false. % notchLineColor - Color of the notch line when notchLine=true. % The default is 'k'. % notchLineStyle - Line style of the notch line when % notchLine=true. The default is ':'. % outlierSize - Size, in square points, of the outlier % marker. The default is 36. % percentile - Percentile limits of the boxes. The default % is [25,75]. All other statistics, such as % those that depend on the IQR (e.g. whisker % extent and notch height), are unaffected by % this parameter. % sampleFontSize - Specify the font size of the sample size % display (if sampleSize=true). The default is % 9. % sampleSize - Specify whether to display the sample size in % the top-right of each box. The default is % false. % scaleWidth - Logical value to specify scaling the width of % each box according to the square root of the % sample size. The default is false. % scatterAlpha - Set the transparency of the scatter markers % (0 is transparent, 1 is opaque) when % showScatter=true. The default is 1. % scatterColor - Scatter marker color for scatter plots of % underlying data (if showScatter=true). The % default is [.5 .5 .5]. % scatterLayer - Set the layer of scatter plots with respect % to the boxes. Can be set to 'top' or % 'bottom'. The default is 'top'. % scatterMarker - Marker used for scatter plots of underlying % data (if showScatter=true). The default is % 'x'. % scatterSize - Size, in square points, of the scatter % markers (if showScatter=true). The default is % 36. % showLegend - Display a legend of the data. The labels can % be set using the 'groupLabels' option. Note % that the legend uses the box color, median % line, or violin fill color to distinguish % legend entries. If these properties do not % differ between boxes then an error will be % returned. % showMean - Logical value determines whether to display % the mean of the data for each box. The % default is false. % showOutliers - Logical value determines whether to display % outliers. The default is true. % showScatter - If set to true, a scatter plot of the % underlying data for each box will be % overlayed on the plot. The data will have a % random x-axis offset with respect to the box % centre. Data that are outliers are not % included. The default is false. % showViolin - If true, a 'violin' [1] kernel density % outline of the data underlying each box will % be plotted. The outline is calculated using % the IOSR.STATISTICS.KERNELDENSITY function. % The default is false. % style - Determine whether to show additional x-axis % labels for the data. If set to 'hierarchy', % additional hierarchical x-axis labels will be % added for the plot. The labels can be set % using the 'groupLabels' option. If set to % 'normal' (default) no additional labels will % be plotted. % symbolColor - Outlier marker color. The default is % 'auto' (see boxColor). % symbolMarker - Marker used to denote outliers. The default % is 'o'. % theme - Specify a display theme to change multiple % display properties. The options are: % 'colorall' : colored boxes (with 0.4 % alpha), lines, and % scatter markers (median % line and mean marker are % black) % 'colorlines' : clear boxes, black mean % markers, colored lines % and scatter markers % 'colorboxes' : colored boxes, black % lines and markers, gray % scatter markers % 'default' : clear boxes, colored % median lines and markers, % gray scatter markers % Use the 'themeColors' property to specify the % color(s) that are used. % themeColors Colors used when creating the theme. The % default is 'auto' (see boxColor). % violinBins - If 'showViolin' is true, this specifes the % bins used to calculate the kernel density. % See IOSR.STATISTICS.KERNELDENSITY. % violinBinWidth - If 'showViolin' is true, this specifes the % bin width used to calculate the kernel % density. See IOSR.STATISTICS.KERNELDENSITY. % violinColor - Fill color for the violins. The default is % 'none' (see boxColor). % violinKernel - If 'showViolin' is true, this specifes the % kernel used to calculate the kernel density. % See IOSR.STATISTICS.KERNELDENSITY. % xSeparator - Logical value that when true adds a separator % line between x groups. The default is false. % xSpacing - Determine the x-axis spacing of boxes. By % default ('x'), the data in x are used to % determine the position of boxes on the x-axis % (when x is numeric). Alternativley, when set % to 'equal', boxes are equally-spaced, but the % data in x are used to label the axis; the % x-axis ticks are at 1:LENGTH(X). % handles - Structure containing handles to the various % objects that constitute the plot. The % fields/handles are: % 'axes' : the parent axes of % the box plot % 'fig' : the parent figure of % the box plot % 'addPrctiles' : chart line objects % for each additional % percentile marker % 'addPrctilesTxt' : text objects for each % additional percentile % marker % 'box' : patch objects for % each box % 'medianLines' : line objects for each % median line % 'means' : chart line objects % for each mean marker % 'notchLowerLines' : line objects for each % lower notch line % 'notchUpperLines' : line objects for each % upper notch line % 'upperWhiskers' : line objects for each % upper whisker line % 'lowerWhiskers' : line objects for each % lower whisker line % 'upperWhiskerTips': line objects for each % upper whisker tip % 'lowerWhiskerTips': line objects for each % lower whisker line % 'outliers' : scatter objects for % each set of outliers % 'scatters' : scatter objects for % each scatter overlay % 'samplesTxt' : text objects for each % sample size display % 'groupsTxt' : text objects for each % group label % 'xseps' : line objects for each % x separator % 'legend' : the legend object % % These properties can be referenced using dot notation - e.g. H.BOXCOLOR % where H is an instance of the BOXPLOT object - or using the SET and GET % methods - e.g. GET(H,'BOXCOLOR'). Both methods are case-insensitive. % % Note that some handles will be empty unless the associated option is % turned on. % % Read-only properties: % x - The x data. % y - The y data. % weights - Array giving the weights for the data in y. % The array must be the same size as y. This % option may be specified in the constructor. % statistics - Structure containing the statistics used % for the box plot. With the exception of % 'outliers', 'outliers_IX', 'addPrctiles', and % 'percentile' noted below, each field contains % a 1-by-P-by-G-by-I-by-J... numeric array of % values (identical to that returned by % IOSR.STATISTICS.QUANTILE). The fields are: % 'addPrctiles' : an A-by-P-by-G... array % containing the % additional percentile % values specified by the % 'addPrctiles' property, % where A is the number % of percentiles in the % property % 'percentile' : the percentile limits % specified by the % 'percentile' property % 'median' : the median values % 'N' : the sample size % 'PL' : the lower percentiles % specified by the % 'percentile' property, % and the lower limit of % the boxes (default is % 25th percentile) % 'PU' : the upper percentiles % specified by the % 'percentile' property, % and the upper limit of % the boxes (default is % 75th percentile) % 'Q1' : the 25th percentiles % 'Q3' : the 75th percentiles % 'IQR' : the inter-quartile % ranges % 'mean' : the mean values % 'min' : the minimum values % (excl. outliers) % 'max' : the maximum values % (excl. outliers) % 'notch_u' : the upper notch values % 'notch_l' : the lower notch values % 'outliers' : a 1-by-P-by-G cell % array of outlier values % 'outliers_IX' : a logical array, the % same size as Y, with % true values indicating % outliers % 'std' : the standard deviations % % In addition to the above specifications, some options can be specified % for each group. Parameters should be specified as a cell array of size % G-by-I-by-J... . These options are: 'boxColor', 'lineColor', % 'lineStyle', 'meanColor', 'meanMarker, 'medianColor', 'notchLineColor', % 'notchLineStyle', 'scatterMarker', 'scatterColor', 'symbolColor', % 'symbolMarker', 'themeColor', and 'violinColor'. % % As noted above, colors may be specified as a colormap function handle % (e.g. @gray for grayscale colors). The function handle can refer to one % of the built-in colormap functions, or any other function capable of % generating valid colormaps. If the specified colormap is one of the % built-in colormaps 'hot' (@hot), 'gray' (@gray), 'bone' (@bone), % 'copper' (@copper), or 'pink' (@pink), then boxPlot will take steps to % try to restrict the range of colors. Specifically, for fills (e.g. % 'boxColor'), the colors will be restricted such that the minimum % luminance is 0.33 (in order to not mask dark lines). For lines, the % colors will be restricted such that the maximum luminance is 0.66 (in % order that lines are not masked by the white background). % % IOSR.STATISTICS.BOXPLOT methods: % boxPlot - Create the box plot. % % See also IOSR.STATISTICS.TAB2BOX, IOSR.STATISTICS.QUANTILE, COLORMAP, % IOSR.STATISTICS.FUNCTIONALSPREADPLOT, % IOSR.STATISTICS.FUNCTIONALBOXPLOT, IOSR.STATISTICS.KERNELDENSITY. % % References % % [1] Hintze, Jerry L.; Nelson, Ray D. (1998). "Violin Plots: A Box % Plot-Density Trace Synergism". The American Statistician. 52 (2): % 181?4. % Some modifications. Javier G. Orlandi % Copyright 2016 University of Surrey. properties (AbortSet) addPrctiles = [] % Additional percentiles to plot. addPrctilesColors % Colors for the additional percentile markers. addPrctilesLabels = {} % Labels for additional percentiles. addPrctilesMarkers = {} % Markers for additional percentiles. addPrctilesSize % Size of the additional percentile markers. addPrctilesTxtSize % Size of the additional percentile labels text. boxAlpha = 1 % The transparency of the boxes. boxColor = 'none' % Fill color of the boxes. boxWidth = 'auto' % The width of the boxes. groupLabelFontSize = 9 % The font size of the group labels. groupLabelHeight = 'auto' % The height of the area reserved for group labels. groupLabels = [] % Labels used to label the boxes for each x-group. groupWidth = 0.75 % The proportion of the x-axis interval across which each x-group of boxes should be spread. limit = '1.5IQR' % Mode indicating the limits that define outliers. lineColor = 'k' % Color of the box outlines and whiskers. lineStyle = '-' % Style of the whisker lines. lineWidth = 1 % Width, in points, of the box outline, whisker lines, notch line, and outlier marker edges. meanColor = 'auto' % Color of the mean marker. meanMarker = '+' % Marker used for the mean. meanSize = 6 % Size, in point units, of the mean markers. medianColor = 'auto' % Color of the median line. method = 'R-8' % The method used to calculate the quantiles. notch = false % Whether the box should have a notch. notchDepth = 0.4 % Depth of the notch as a proportion of half the box width. notchLineColor = 'k' % Color of the notch line. notchLineStyle = ':' % Line style of the notch line. notchLine = false % Whether to draw a horizontal line in the box at the extremes of the notch. outlierSize = 6^2 % Size, in square points, of the outlier markers. percentile = [25,75] % Percentile limits of the box. sampleFontSize = 9 % Specify the font size of the sample size display. sampleSize = false % Whether to display the sample size in the top-right of each box. scaleWidth = false % Scale the width of each box according to the square root of the sample size. scatterAlpha = 1 % The transparency of the scatter markers. scatterColor = [.5 .5 .5] % Scatter marker color for scatter plots of underlying data. scatterLayer = 'top' % The layer of scatter plots with respect to the boxes. scatterMarker = 'x' % Marker used for scatter plots of underlying data. scatterSize = 6^2 % Size, in square points, of the scatter markers. showLegend = false % Draw a legend. showMean = false % Display the mean of the data for each box. showOutliers = true % Display outliers. showScatter = false % Display a scatter plot of the underlying data for each box. showViolin = false % Display a violin (kernel density) plot for the underlying data. style = 'normal' % Determine whether to show additional x-axis labels for the data. symbolColor = 'auto' % Outlier marker color. symbolMarker = 'o' % Marker used to denote outliers. theme = 'default' % Control a range of display properties. themeColors = 'auto' % Colors used when creating the theme. violinBins = 'auto' % The bins used to calculate the violins. violinBinWidth = 'auto' % The width of the bins used to calculate the violins. violinAlpha = 1 % Alpha of the violins. violinColor = 'none' % Color of the violins. violinWidth = 'auto' % The width of the violins. violinKernel = 'normal' % The violin kernel used to calculate the kernel density. xSeparator = false % Add a separator line between x groups. xSpacing = 'x' % Determine the x-axis spacing of boxes. rescaleGroups = false; % Will rescale groups based on their real # of elements end properties (Access = private) diffx % smallest difference of x axis data groupXticks % x ticks for every box groupDims % dimensions of data for each x group groupRange % scale value range of each x group xlabels % labels for the x axis xticks % x-axis tick points groupSizes % In case a rescaling is needed end properties (SetAccess = private, Dependent, Hidden) addPrctilesHandle = [] % Chart line objects objects for the additional percentiles markers. addPrctilesTxtHandle = [] % Text objects for each additional percentile label. boxHandles = [] % Patch objects for each box. mLineHandles = [] % Line objects for each median line. meanHandles = [] % Chart line objects for each mean marker. notchLowerLineHandles = [] % Line objects for each lower notch line. notchUpperLineHandles = [] % Line objects for each upper notch line. upperWhiskerHandles = [] % Line objects for each upper whisker line. lowerWhiskerHandles = [] % Line objects for each lower whisker line. upperWhiskerTipHandles = [] % Line objects for each upper whisker tip. lowerWhiskerTipHandles = [] % Line objects for each lower whisker tip. outliersHandles = [] % Scatter objects for each set of outliers. scatterHandles = [] % Scatter objects for each scatter overlay. sampleTxtHandes = [] % Text objects for each sample size display. groupTxtHandes = [] % Text objects for each group label. xsepHandles = [] % Line objects for each x separator. legendHandle = [] % Legend object. end methods function obj = boxPlot(varargin) % BOXPLOT Draw a box plot. % % IOSR.STATISTICS.BOXPLOT(Y) produces a box plot of the data in % Y. If Y is a vector, there is just one box. If Y is a matrix, % there is one box per column. If Y is an % N-by-P-by-G-by-I-by-J... array then G*I*J*... boxes are plotted % hierarchically for each column P; i.e. J boxes are plotted for % each index of I, and J*I boxes are plotted for each index of G. % For each box, the central mark is the median of the % data/column, the edges of the box are the 25th and 75th % percentiles, the whiskers extend to the most extreme data % points not considered outliers, and outliers are plotted % individually. NaNs are excluded from the data. % % Tabular data can be arranged into the appropriate format using % the IOSR.STATISTICS.TAB2BOX function. % % IOSR.STATISTICS.BOXPLOT(X,Y) specifies the x-axis values for % each box. X should be a vector, with as many elements as Y has % columns. The default is 1:SIZE(Y,2). % % IOSR.STATISTICS.BOXPLOT(...,'PARAMETER',VALUE) allows the % plotting options to be specified when the plot is constructed. % % IOSR.STATISTICS.BOXPLOT(AX,...) creates the box plot in the % axes specified by AX. % % Examples % % Example 1: Basic grouped box plot with legend % % y = randn(50,3,3); % x = [1 2 3.5]; % y(1:25) = NaN; % % figure; % h = iosr.statistics.boxPlot(x,y,... % 'symbolColor','k',... % 'medianColor','k',... % 'symbolMarker',{'+','o','d'},... % 'boxcolor',{[1 0 0]; [0 1 0]; [0 0 1]},... % 'groupLabels',{'y1','y2','y3'},... % 'showLegend',true); % box on % % Example 2: Grouped box plot with overlayed data % % figure; % iosr.statistics.boxPlot(x,y,... % 'symbolColor','k',... % 'medianColor','k',... % 'symbolMarker',{'+','o','d'},... % 'boxcolor','auto',... % 'showScatter',true); % box on % % Example 3: Grouped box plot with displayed sample sizes % and variable widths % % figure; % iosr.statistics.boxPlot(x,y,... % 'medianColor','k',... % 'symbolMarker',{'+','o','d'},... % 'boxcolor','auto',... % 'sampleSize',true,... % 'scaleWidth',true); % box on % % Example 4: Grouped notched box plot with x separators and % hierarchical labels % % figure; % iosr.statistics.boxPlot({'A','B','C'},y,... % 'notch',true,... % 'medianColor','k',... % 'symbolMarker',{'+','o','d'},... % 'boxcolor','auto',... % 'style','hierarchy',... % 'xSeparator',true,... % 'groupLabels',{{'Group 1','Group 2','Group 3'}}); % box on % % Example 5: Box plot with legend labels from data % % % load data % % (requires Statistics or Machine Learning Toolbox) % load carbig % % % arrange data % [y,x,g] = iosr.statistics.tab2box(Cylinders,MPG,when); % % % sort % IX = [1 3 2]; % order % g = g{1}(IX); % y = y(:,:,IX); % % % plot % figure % h = iosr.statistics.boxPlot(x,y,... % 'symbolColor','k','medianColor','k','symbolMarker','+',... % 'boxcolor',{[1 1 1],[.75 .75 .75],[.5 .5 .5]},... % 'scalewidth',true,'xseparator',true,... % 'groupLabels',g,'showLegend',true); % box on % title('MPG by number of cylinders and period') % xlabel('Number of cylinders') % ylabel('MPG') % % Example 6: Box plot calculated from weighted quantiles % % % load data % load carbig % % % random weights % weights = rand(size(MPG)); % % % arrange data % [y,x,g] = iosr.statistics.tab2box(Cylinders,MPG,when); % weights_boxed = iosr.statistics.tab2box(Cylinders,weights,when); % % % plot % figure % h = iosr.statistics.boxPlot(x,y,'weights',weights_boxed); % % Example 7: Draw a violin plot % y = randn(50,3,3); % x = [1 2 3.5]; % y(1:25) = NaN; % figure('color','w'); % h2 = iosr.statistics.boxPlot(x,y, 'showViolin', true, 'boxWidth', 0.025, 'showOutliers', false); % box on if nargin > 0 %% check for deprecated properties % check for scatter option scatterIX = strcmpi('scatter',varargin); if any(scatterIX) warning('The ''scatter'' property is deprecated. Use ''showScatter'' instead.'); varargin{scatterIX} = 'showScatter'; end % check for mean option meanIX = strcmpi('mean',varargin); if any(meanIX) warning('The ''mean'' property is deprecated. Use ''showMean'' instead.'); varargin{meanIX} = 'showMean'; end %% set x, y, and dims % check for input data start = obj.getXY(varargin{:}); obj.groupDims = obj.ydims(3:end); % create initial theme obj.createTheme(); % check weights obj.checkWeights(); % set properties from varargin obj.setProperties(start,nargin,varargin); % remove NaN columns obj.removeNaN(); % set x axis properties and labels obj.setXprops(); %% statistics % calculate statistics obj.calculateStats(); %% draw % set handles obj.parseAxesHandle(varargin{:}); % draw the box plot obj.draw('all'); % add listener to ylim so that some properties can be redrawn addlistener(handle(obj.handles.axes),'YLim','PostSet',... @(hProp,eventData) obj.eventHandler(hProp,eventData)); end end %% dependent handle getters % these are all deprecated and the handles moved to the obj.handles struct function val = get.addPrctilesHandle(obj) warning('iosr:boxPlot:deprecatedProp','''obj.addPrctilesHandle'' is deprecated. Use ''obj.handles.addPrctiles'' instead.'); val = obj.handles.addPrctiles; end function val = get.addPrctilesTxtHandle(obj) warning('iosr:boxPlot:deprecatedProp','''obj.addPrctilesTxtHandle'' is deprecated. Use ''obj.handles.addPrctilesTxt'' instead.'); val = obj.handles.addPrctilesTxt; end function val = get.boxHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.boxHandles'' is deprecated. Use ''obj.handles.box'' instead.'); val = obj.handles.box; end function val = get.mLineHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.mLineHandles'' is deprecated. Use ''obj.handles.medianLines'' instead.'); val = obj.handles.medianLines; end function val = get.meanHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.meanHandles'' is deprecated. Use ''obj.handles.means'' instead.'); val = obj.handles.means; end function val = get.notchLowerLineHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.notchLowerLineHandles'' is deprecated. Use ''obj.handles.notchLowerLines'' instead.'); val = obj.handles.notchLowerLines; end function val = get.notchUpperLineHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.notchUpperLineHandles'' is deprecated. Use ''obj.handles.notchUpperLines'' instead.'); val = obj.handles.notchUpperLines; end function val = get.upperWhiskerHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.upperWhiskerHandles'' is deprecated. Use ''obj.handles.upperWhiskers'' instead.'); val = obj.handles.upperWhiskers; end function val = get.lowerWhiskerHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.lowerWhiskerHandles'' is deprecated. Use ''obj.handles.lowerWhiskers'' instead.'); val = obj.handles.lowerWhiskers; end function val = get.upperWhiskerTipHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.upperWhiskerTipHandles'' is deprecated. Use ''obj.handles.upperWhiskerTips'' instead.'); val = obj.handles.upperWhiskerTips; end function val = get.lowerWhiskerTipHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.lowerWhiskerTipHandles'' is deprecated. Use ''obj.handles.lowerWhiskerTips'' instead.'); val = obj.handles.lowerWhiskerTips; end function val = get.outliersHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.outliersHandles'' is deprecated. Use ''obj.handles.outliers'' instead.'); val = obj.handles.outliers; end function val = get.scatterHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.scatterHandles'' is deprecated. Use ''obj.handles.scatters'' instead.'); val = obj.handles.scatters; end function val = get.sampleTxtHandes(obj) warning('iosr:boxPlot:deprecatedProp','''obj.sampleTxtHandes'' is deprecated. Use ''obj.handles.samplesTxt'' instead.'); val = obj.handles.samplesTxt; end function val = get.groupTxtHandes(obj) warning('iosr:boxPlot:deprecatedProp','''obj.groupTxtHandes'' is deprecated. Use ''obj.handles.groupsTxt'' instead.'); val = obj.handles.groupsTxt; end function val = get.xsepHandles(obj) warning('iosr:boxPlot:deprecatedProp','''obj.xsepHandles'' is deprecated. Use ''obj.handles.xseps'' instead.'); val = obj.handles.xseps; end function val = get.legendHandle(obj) warning('iosr:boxPlot:deprecatedProp','''obj.legendHandle'' is deprecated. Use ''obj.handles.legend'' instead.'); val = obj.handles.legend; end %% accessor functions % set/get additional percentiles function val = get.addPrctiles(obj) val = obj.addPrctiles; end function set.addPrctiles(obj,val) if ~isempty(val) assert(isvector(val) && isnumeric(val), 'iosr:boxPlot:addPrctilesVector', '''ADDPRCTILES'' should be a numeric vector') assert(all(val<=100) && all(val>=0), 'iosr:boxPlot:addPrctilesRange', '''ADDPRCTILES'' values should be in the interval [0,100].') end obj.addPrctiles = val; obj.calculateStats(); obj.draw('addPrctiles'); end % set/get additional percentiles colors function val = get.addPrctilesColors(obj) val = obj.checkAddPrctileProp(obj.addPrctilesColors,{'k'}); end function set.addPrctilesColors(obj,val) if ~isempty(val) assert(iscell(val), 'iosr:boxPlot:addPrctilesColors', '''ADDPRCTILESCOLORS'' should be a cell array') end obj.addPrctilesColors = val; obj.draw(); end % set/get additional percentiles labels function val = get.addPrctilesLabels(obj) val = obj.checkAddPrctileProp(obj.addPrctilesLabels,{''}); end function set.addPrctilesLabels(obj,val) if ~isempty(val) assert(iscellstr(val), 'iosr:boxPlot:addPrctilesLabels', '''ADDPRCTILESLABELS'' should be a cell array of strings') end obj.addPrctilesLabels = val; obj.draw(); end % set/get additional percentiles markers function val = get.addPrctilesMarkers(obj) val = obj.checkAddPrctileProp(obj.addPrctilesMarkers,{'*'}); end function set.addPrctilesMarkers(obj,val) if ~isempty(val) assert(iscellstr(val), 'iosr:boxPlot:addPrctilesMarkers', '''ADDPRCTILESMARKERS'' should be a cell array of strings') end obj.addPrctilesMarkers = val; obj.draw(); end % set/get additional percentiles marker size function val = get.addPrctilesSize(obj) val = obj.checkAddPrctileProp(obj.addPrctilesSize,6); end function set.addPrctilesSize(obj,val) if ~isempty(val) assert(isnumeric(val) && isvector(val), 'iosr:boxPlot:addPrctilesSize', '''ADDPRCTILESSIZE'' should be a numeric vector') end obj.addPrctilesSize = val; obj.draw(); end % set additional percentiles label font size function set.addPrctilesTxtSize(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:addPrctilesTxtSize', '''ADDPRCTILESTXTSIZE'' should be a numeric scalar') obj.addPrctilesTxtSize = val; obj.draw(); end % set box alpha function set.boxAlpha(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:boxAlpha', '''BOXALPHA'' must be a numeric scalar') obj.boxAlpha = val; obj.draw('legend'); end % set/get box color function val = get.boxColor(obj) val = obj.checkColor(obj.boxColor,'fill'); val = obj.groupOption(val,'boxColor'); end function set.boxColor(obj,val) if(iscell(obj.boxColor) && size(obj.boxColor, 1) == size(val, 1) && size(val, 2) == 3) for it = 1:size(obj.boxColor, 1) obj.boxColor{it} = val(it, :); obj.draw('legend'); end else obj.boxColor = val; obj.draw('legend'); end end % set box width function set.boxWidth(obj,val) assert((isnumeric(val) && isscalar(val)) || strcmpi(val,'auto'), 'iosr:boxPlot:boxWidth', '''BOXWIDTH'' must be a numeric scalar or ''auto''.') obj.boxWidth = val; obj.draw('all'); end % set/get group labels function val = get.groupLabels(obj) if prod(obj.groupDims)==numel(obj.groupLabels) && numel(obj.groupLabels)>1 val = {obj.groupLabels}; else val = obj.groupLabels; end if ~isempty(val) % use input assert(isvector(val) && iscell(val),... 'iosr:boxPlot:groupLabelsType', ... 'The GROUPLABELS option should be a cell vector'); valSize = cellfun(@length,val); assert(prod(obj.outDims(3:end))==prod(valSize), ... 'iosr:boxPlot:groupLabelsSize', ... ['The GROUPLABELS option should be a cell vector; ' ... 'the Nth element should contain a vector of length SIZE(Y,N+2)']) assert(isequal(obj.outDims(3:end),valSize(1:length(obj.outDims)-2)), ... 'iosr:boxPlot:groupLabelsSize', ... ['The GROUPLABELS option should be a cell vector; ' ... 'the Nth element should contain a vector of length SIZE(Y,N+2)']) else % create placeholder labels val = cell(1,length(obj.groupDims)); c = 1; for m = 1:length(val) val{m} = cell(1,obj.groupDims(m)); for n = 1:obj.groupDims(m) val{m}{n} = sprintf('Data %d',c); c = c+1; end end end end function set.groupLabels(obj,val) obj.groupLabels = val; obj.draw('style','legend'); end % set group label font size function set.groupLabelFontSize(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:groupLabelsFontSize', '''GROUPLABELFONTSIZE'' must be a numeric scalar.') obj.groupLabelFontSize = val; obj.draw(); end % set group label height function set.groupLabelHeight(obj,val) assert(strcmp(val,'auto') || (isnumeric(val) && isscalar(val)), 'iosr:boxPlot:groupLabelHeight', '''GROUPLABELHEIGHT'' must be ''auto'' or a numeric scalar.') obj.groupLabelHeight = val; obj.draw('style'); end % set group width function set.groupWidth(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:groupWidth', '''GROUPWIDTH'' must be a numeric scalar') obj.groupWidth = val; obj.setXprops(); obj.draw('all'); end % set stats limit function set.limit(obj,val) if isnumeric(val) assert(isnumeric(val) && numel(val)==2, 'iosr:boxPlot:limitSize', '''LIMIT'' must be a two-element numeric vector.') assert(all(val<=100) && all(val>=0), 'iosr:boxPlot:limitRange', '''LIMIT'' values should be in the interval [0,100].') if all(val<1) warning('iosr:boxPlot:limitRange','''LIMIT'' values should be in the interval [0,100].') end else assert(ischar(val), 'iosr:boxPlot:limitType', '''LIMIT'' must be a char or numeric array.') assert(any(strcmpi(val,{'1.5IQR','3IQR','none'})), 'iosr:boxPlot:limitUnknownOption', '''LIMIT'' parameter not recognised.') end obj.limit = val; obj.calculateStats(); obj.draw('whiskers','outliers','scatter'); end % get/set line colors function val = get.lineColor(obj) val = obj.checkColor(obj.lineColor,'line'); val = obj.groupOption(val,'lineColor'); end function set.lineColor(obj,val) obj.lineColor = val; obj.draw('legend'); end % get/set line style function val = get.lineStyle(obj) val = obj.groupOption(obj.lineStyle,'lineStyle'); end function set.lineStyle(obj,val) assert(ischar(val) || iscellstr(val), 'iosr:boxPlot:lineStyle', '''LINESTYLE'' must be a char array or cell array of strings.') obj.lineStyle = val; obj.draw('legend'); end % set line width function set.lineWidth(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:lineWidth', '''LINEWIDTH'' must be a numeric scalar.') obj.lineWidth = val; obj.draw('all'); end % get/set mean color function val = get.meanColor(obj) val = obj.checkColor(obj.meanColor,'line'); val = obj.groupOption(val,'meanColor'); end function set.meanColor(obj,val) obj.meanColor = val; obj.draw(); end % get/set mean marker function val = get.meanMarker(obj) val = obj.groupOption(obj.meanMarker,'meanMarker'); end function set.meanMarker(obj,val) assert(ischar(val) || iscellstr(val), 'iosr:boxPlot:meanMarker', '''MEANMARKER'' must be a char array or cell array of strings.') obj.meanMarker = val; obj.draw(); end % set mean size function set.meanSize(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:meanSize', '''MEANSIZE'' must be a numeric scalar.') obj.meanSize = val; obj.draw(); end % get/set median color function val = get.medianColor(obj) val = obj.checkColor(obj.medianColor,'line'); val = obj.groupOption(val,'medianColor'); end function set.medianColor(obj,val) obj.medianColor = val; obj.draw('legend'); end % set stats method function set.method(obj,val) assert(ischar(val), 'iosr:boxPlot:method', '''METHOD'' must be a char array.') obj.method = val; obj.calculateStats(); obj.draw('all'); end % set notch function set.notch(obj,val) assert(islogical(val) && numel(val)==1, 'iosr:boxPlot:notch', '''NOTCH'' must be logical.') obj.notch = val; obj.draw('boxes'); end % set notch depth function set.notchDepth(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:notchDepth', '''NOTCHDEPTH'' must be a numeric scalar') obj.notchDepth = val; obj.draw('boxes'); end % get/set notch line color function val = get.notchLineColor(obj) val = obj.checkColor(obj.notchLineColor,'line'); val = obj.groupOption(val,'notchLineColor'); end function set.notchLineColor(obj,val) obj.notchLineColor = val; obj.draw(); end % get/set notch line style function val = get.notchLineStyle(obj) val = obj.groupOption(obj.notchLineStyle,'notchLineStyle'); end function set.notchLineStyle(obj,val) assert(ischar(val) || iscellstr(val), 'iosr:boxPlot:notchLineStyle', '''NOTCHLINESTYLE'' must be a char array or cell array of strings.') obj.notchLineStyle = val; obj.draw(); end % set notch line function set.notchLine(obj,val) assert(islogical(val) && numel(val)==1, 'iosr:boxPlot:notchLine', '''NOTCHLINE'' must be logical.') obj.notchLine = val; obj.draw('boxes'); end % set outlier size function set.outlierSize(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:outlierSize', '''OUTLIERSIZE'' must be a numeric scalar.') obj.outlierSize = val; obj.draw(); end % set percentile function set.percentile(obj,val) assert(isnumeric(val) && numel(val)==2, 'iosr:boxPlot:percentileSize', '''PERCENTILE'' must be a two-element numeric vector.') assert(all(val<=100) && all(val>=0), 'iosr:boxPlot:percentileRange', '''PERCENTILE'' values should be in the interval [0,100].') if all(val<1) warning('iosr:boxPlot:percentileRange', '''PERCENTILE'' values should be in the interval [0,100].') end obj.percentile = val; obj.calculateStats(); obj.draw('all'); end % set sample font size function set.sampleFontSize(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:sampleFontSize', '''SAMPLEFONTSIZE'' must be a numeric scalar.') obj.sampleFontSize = val; obj.draw(); end % set sample size option function set.sampleSize(obj,val) assert(islogical(val) && numel(val)==1, 'iosr:boxPlot:sampleSize', '''SAMPLESIZE'' must be logical.') obj.sampleSize = val; obj.draw('boxes'); end % set scale width option function set.scaleWidth(obj,val) assert(islogical(val) && numel(val)==1, 'iosr:boxPlot:scaleWidth', '''SCALEWIDTH'' must be logical.') obj.scaleWidth = val; obj.draw('boxes','scatter','outliers'); end % set legend function set.showLegend(obj,val) assert(islogical(val) && numel(val)==1, 'iosr:boxPlot:showLegend', '''SHOWLEGEND'' must be logical.') obj.showLegend = val; obj.draw('legend'); end % set show mean option function set.showMean(obj,val) assert(islogical(val) && numel(val)==1, 'iosr:boxPlot:showMean', '''SHOWMEAN'' must be logical.') obj.showMean = val; obj.draw('means'); end % set show outliers option function set.showOutliers(obj,val) assert(islogical(val) && numel(val)==1, 'iosr:boxPlot:showOutliers', '''SHOWOUTLIERS'' must be logical.') obj.showOutliers = val; obj.draw('outliers'); end % set show scatter option function set.showScatter(obj,val) assert(islogical(val) && numel(val)==1, 'iosr:boxPlot:showScatter', '''SHOWSCATTER'' must be logical.') obj.showScatter = val; obj.draw('scatter'); end % set scatter alpha function set.scatterAlpha(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:scatterAlpha', '''SCATTERALPHA'' must be a numeric scalar') obj.scatterAlpha = val; obj.draw(); end % get/set scatter color function val = get.scatterColor(obj) val = obj.checkColor(obj.scatterColor,'line'); val = obj.groupOption(val,'scatterColor'); end function set.scatterColor(obj,val) obj.scatterColor = val; obj.draw(); end % set scatter layer function set.scatterLayer(obj,val) assert(ischar(val), 'iosr:boxPlot:scatterLayer', '''SCATTERLAYER'' must be a char array.') obj.scatterLayer = val; obj.draw(); end % get/set scatter marker function val = get.scatterMarker(obj) val = obj.groupOption(obj.scatterMarker,'scatterMarker'); end function set.scatterMarker(obj,val) assert(ischar(val) || iscellstr(val), 'iosr:boxPlot:scatterMarker', '''SCATTERMARKER'' must be a char array or cell array of strings.') obj.scatterMarker = val; obj.draw(); end % set scatter marker size function set.scatterSize(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:scatterSize', '''SCATTERSIZE'' must be a numeric scalar.') obj.scatterSize = val; obj.draw(); end % violin function set.showViolin(obj,val) assert(islogical(val) && isscalar(val), 'iosr:boxPlot:showViolin', '''VIOLIN'' must be logical.') obj.showViolin = val; obj.draw('violin','whiskers'); end % set style function set.style(obj,val) assert(ischar(val), 'iosr:boxPlot:styleType', '''STYLE'' must be a char array.') assert(any(strcmpi(val,{'normal','hierarchy'})), 'iosr:boxPlot:styleOption', '''STYLE'' must be either ''normal'' or ''hierarchy''.') obj.style = val; obj.draw('style'); end % get/set outlier symbol color function val = get.symbolColor(obj) val = obj.checkColor(obj.symbolColor,'line'); val = obj.groupOption(val,'symbolColor'); end function set.symbolColor(obj,val) obj.symbolColor = val; obj.draw(); end % get/set outlier symbol marker function val = get.symbolMarker(obj) val = obj.groupOption(obj.symbolMarker,'symbolMarker'); end function set.symbolMarker(obj,val) assert(ischar(val) || iscellstr(val), 'iosr:boxPlot:symbolMarker', '''SYMBOLMARKER'' must be a char array or cell array of strings.') obj.symbolMarker = val; obj.draw(); end % set theme function set.theme(obj,val) assert(ischar(val), 'iosr:boxPlot:theme', '''THEME'' must be a char array.') obj.theme = val; obj.createTheme(); obj.draw('legend'); end % theme color function val = get.themeColors(obj) switch lower(obj.theme) case {'graylines', 'colorlines', 'default'} val = obj.checkColor(obj.themeColors,'lines'); case {'grayboxes', 'colorall', 'colorboxes'} val = obj.checkColor(obj.themeColors,'fill'); end val = obj.groupOption(val,'themeColors'); end function set.themeColors(obj,val) obj.themeColors = val; obj.createTheme(); obj.draw('all'); end % set violin props function set.violinBins(obj,val) assert(isnumeric(val) || strcmpi(val,'auto'), 'iosr:boxPlot:violinBins', '''VIOLINBINS'' must be numeric or ''auto''.') obj.violinBins = val; obj.draw('violin','whiskers'); end function set.violinBinWidth(obj,val) assert((isnumeric(val) && isscalar(val)) || strcmpi(val,'auto'), 'iosr:boxPlot:violinBinWidth', '''VIOLINBINWIDTH'' must be a numeric scalar or ''auto''.') obj.violinBinWidth = val; obj.draw('violin','whiskers'); end function set.violinKernel(obj,val) assert(ischar(val), 'iosr:boxPlot:violinKernel', '''VIOLINKERNEL'' must be a string.') obj.violinKernel = val; obj.draw('violin','whiskers'); end function set.violinAlpha(obj,val) assert(isnumeric(val) && isscalar(val), 'iosr:boxPlot:violinAlpha', '''VIOLINALPHA'' must be a numeric scalar') obj.violinAlpha = val; obj.draw('legend'); end function val = get.violinColor(obj) val = obj.checkColor(obj.violinColor,'fill'); val = obj.groupOption(val,'violinColor'); end function set.violinColor(obj,val) obj.violinColor = val; obj.draw('legend'); end function set.violinWidth(obj,val) assert((isnumeric(val) && isscalar(val)) || strcmpi(val,'auto'), 'iosr:boxPlot:violinWidth', '''VIOLINWIDTH'' must be a numeric scalar or ''auto''.') obj.violinWidth = val; obj.draw('violin','whiskers'); end % set x separator option function set.xSeparator(obj,val) assert(islogical(val) && isscalar(val), 'iosr:boxPlot:xSeparator', '''XSEPARATOR'' must be logical.') obj.xSeparator = val; obj.draw('xsep'); end % set x spacing option function set.xSpacing(obj,val) assert(any(strcmpi(val,{'x','equal'})), 'iosr:boxPlot:xSpacing', '''XSPACING'' parameter not recognised.') obj.xSpacing = val; obj.setXprops(); obj.draw('all'); end end methods (Access = protected) function draw(obj,varargin) %DRAW main draw function if isfield(obj.handles,'axes') subidx = cell(1,length(obj.outDims)); hold on; for n = 1:prod(obj.outDims) % get indices [subidx{:}] = ind2sub(obj.outDims, n); subidxAll = [{':'} subidx(2:end)]; gidx = subidx(3:end); % determine where to put tick marks for group labels obj.groupXticks(subidx{:}) = obj.xticks(subidx{2})+obj.getOffset(subidx); % plot individual components if any(strcmpi('violin',varargin)) || any(strcmpi('all',varargin)) obj.drawViolin(subidx,subidxAll); end if any(strcmpi('boxes',varargin)) || any(strcmpi('all',varargin)) obj.drawBox(subidx,subidxAll); end if any(strcmpi('whiskers',varargin)) || any(strcmpi('all',varargin)) obj.drawWhiskers(subidx); end if any(strcmpi('outliers',varargin)) || any(strcmpi('all',varargin)) obj.drawOutliers(subidx); end if any(strcmpi('scatter',varargin)) || any(strcmpi('all',varargin)) obj.drawScatter(subidx,gidx,subidxAll); end if any(strcmpi('means',varargin)) || any(strcmpi('all',varargin)) obj.drawMean(subidx); end if any(strcmpi('addPrctiles',varargin)) || any(strcmpi('all',varargin)) obj.drawAddPrctiles(subidx); end % change graphics options obj.drawGroupGraphics(subidx,gidx); end % style things if any(strcmpi('xsep',varargin)) || any(strcmpi('all',varargin)) obj.drawXsepatator(); end if any(strcmpi('style',varargin)) || any(strcmpi('all',varargin)) obj.drawStyle(); end if any(strcmpi('legend',varargin)) || any(strcmpi('all',varargin)) obj.drawLegend(); end % change graphics options obj.drawGlobalGraphics(); % set layering if isfield(obj.handles,'scatters') uistack(obj.handles.scatters(:),obj.scatterLayer); end if isfield(obj.handles,'samplesTxt') uistack(obj.handles.samplesTxt(:),'top'); end if isfield(obj.handles,'means') uistack(obj.handles.means(:),'top'); end if isfield(obj.handles,'addPrctiles') uistack(obj.handles.addPrctiles(:),'top'); end if isfield(obj.handles,'addPrctilesTxt') uistack(obj.handles.addPrctilesTxt(:),'top'); end hold off; end end function drawViolin(obj,subidx,subidxAll) %DRAWVIOLIN Draw the violin try % to delete the handles first delete(obj.handles.violin(subidx{:})); catch end if obj.showViolin vBinWidth = obj.replaceAuto(... obj.violinBinWidth, ... [], ... 'Unknown ''VIOLINBINWIDTH'' parameter.'); vBins = obj.replaceAuto(... obj.violinBins, ... [], ... 'Unknown ''VIOLINBINS'' parameter.'); [d, xd] = iosr.statistics.kernelDensity(obj.y(subidxAll{:}), vBins, vBinWidth, obj.violinKernel); halfviolinwidth = obj.calcHalfViolinWidth(subidx); d = halfviolinwidth .* (d ./ max(d)); obj.handles.violin(subidx{:}) = patch(... [d; -flipud(d)] + obj.xticks(subidx{2}) + obj.getOffset(subidx), ... [xd; flipud(xd)],'w','Parent',obj.handles.axes); end end function drawBox(obj,subidx,subidxAll) %DRAWBOX draw boxes try % to delete the handles first delete(obj.handles.box(subidx{:})); catch end try delete(obj.handles.medianLines(subidx{:})); catch end try delete(obj.handles.notchUpperLines(subidx{:})); catch end try delete(obj.handles.notchLowerLines(subidx{:})); catch end try delete(obj.handles.samplesTxt(subidx{:})); catch end % determine half box width [~, halfboxwidth] = obj.calcBoxWidths(subidx); % notch depth notchdepth = obj.notchDepth*halfboxwidth; % main nodes X = obj.xticks(subidx{2}) + obj.getOffset(subidx); q1 = obj.statistics.PL(subidx{:}); q3 = obj.statistics.PU(subidx{:}); md = obj.statistics.median(subidx{:}); nu = obj.statistics.notch_u(subidx{:}); nl = obj.statistics.notch_l(subidx{:}); % set nodes if obj.notch % if notch requested Xnodes = [X-halfboxwidth X-halfboxwidth X-halfboxwidth+notchdepth X-halfboxwidth X-halfboxwidth ... X+halfboxwidth X+halfboxwidth X+halfboxwidth-notchdepth X+halfboxwidth X+halfboxwidth]; Ynodes = [q1 nl md nu q3 q3 nu md nl q1]; else % if not Xnodes = [X-halfboxwidth X-halfboxwidth X+halfboxwidth X+halfboxwidth]; Ynodes = [q1 q3 q3 q1]; end % draw box and set props obj.handles.box(subidx{:}) = patch(Xnodes,Ynodes,'w','Parent',obj.handles.axes); if obj.notchLine % if notchLine requested obj.handles.notchLowerLines(subidx{:}) = line([X-halfboxwidth X+halfboxwidth],[nl nl],... 'Parent',obj.handles.axes); obj.handles.notchUpperLines(subidx{:}) = line([X-halfboxwidth X+halfboxwidth],[nu nu],... 'Parent',obj.handles.axes); end if obj.notch % if notch requested % median obj.handles.medianLines(subidx{:}) = line([X-halfboxwidth+notchdepth X+halfboxwidth-notchdepth], [md md]); else % median obj.handles.medianLines(subidx{:}) = line([X-halfboxwidth X+halfboxwidth],[md md]); end set(obj.handles.medianLines(subidx{:}),'linestyle','-'); if obj.sampleSize % if sample size requested [~, halfboxwidth] = obj.calcBoxWidths(subidx); % set x and y offsets xoffset = 0.15*halfboxwidth; yoffset = (xoffset./(max(obj.xticks)-min(obj.xticks))).*(max(obj.y(:))-min(obj.y(:))); gOffset = obj.getOffset(subidx); % make text obj.handles.samplesTxt(subidx{:}) = text(obj.xticks(subidx{2})+gOffset+halfboxwidth-xoffset,... obj.statistics.PU(subidx{:})-yoffset,... num2str(sum(~isnan(obj.y(subidxAll{:})))),... 'horizontalalignment','right','verticalalignment','top'); end end function drawWhiskers(obj,subidx) %DRAWWHISKERS draw the whiskers try % to delete the handles first delete(obj.handles.upperWhiskers(subidx{:})); catch end try delete(obj.handles.lowerWhiskers(subidx{:})); catch end try delete(obj.handles.upperWhiskerTips(subidx{:})); catch end try delete(obj.handles.lowerWhiskerTips(subidx{:})); catch end % vertices X = obj.xticks(subidx{2}) + obj.getOffset(subidx); [~, halfboxwidth] = calcBoxWidths(obj,subidx); % LQ obj.handles.lowerWhiskers(subidx{:}) = ... line([X X],[obj.statistics.min(subidx{:}) obj.statistics.PL(subidx{:})],'Parent',obj.handles.axes); % UQ obj.handles.upperWhiskers(subidx{:}) = ... line([X X],[obj.statistics.max(subidx{:}) obj.statistics.PU(subidx{:})],'Parent',obj.handles.axes); if ~obj.showViolin % whisker tips obj.handles.lowerWhiskerTips(subidx{:}) = ... line([X-0.5*halfboxwidth X+0.5*halfboxwidth],[obj.statistics.min(subidx{:}) obj.statistics.min(subidx{:})],... 'linestyle','-','Parent',obj.handles.axes); obj.handles.upperWhiskerTips(subidx{:}) = ... line([X-0.5*halfboxwidth X+0.5*halfboxwidth],[obj.statistics.max(subidx{:}) obj.statistics.max(subidx{:})],... 'linestyle','-','Parent',obj.handles.axes); end end function drawOutliers(obj,subidx) %DRAWOUTLIERS draw the outliers try % to delete the handles first delete(obj.handles.outliers(subidx{:})); catch end if obj.showOutliers if ~isempty(obj.statistics.outliers{subidx{:}}) % don't bother if no data % initial plotting vertices [~, halfboxwidth] = obj.calcBoxWidths(subidx); X = obj.xticks(subidx{2}) + obj.getOffset(subidx); % add a random x offset based on data distribution % use full data to calculate offset subidxAll = subidx; subidxAll{1} = ':'; yScatter = obj.y(subidxAll{':'}); randOffset = obj.xOffset(yScatter); ix = obj.statistics.outliers_IX(subidxAll{:}); ix = ix(~isnan(yScatter)); randOffset = randOffset(ix); xScatter = X + (0.8.*halfboxwidth.*randOffset); obj.handles.outliers(subidx{:}) = scatter(xScatter,obj.statistics.outliers{subidx{:}},... 'Parent',obj.handles.axes); end end end function drawScatter(obj,subidx,gidx,subidxAll) %DRAWSCATTER draw the scatter overlay try % to delete the handles first delete(obj.handles.scatters(subidx{:})); catch end if obj.showScatter % get non-outliers IX = ~obj.statistics.outliers_IX(subidxAll{:}); subidx4 = [{IX} subidx(2:end)]; % get data X = obj.xticks(subidx{2}) + obj.getOffset(subidx); Y = obj.y(subidx4{:}); % calculate x offsets [~,halfboxwidth] = calcBoxWidths(obj,subidx); if obj.showViolin halfboxwidth = max(halfboxwidth, obj.calcHalfViolinWidth(subidx)); end xScatter = X + (0.8.*halfboxwidth.*obj.xOffset(Y)); % plot yScatter = Y(~isnan(Y)); obj.handles.scatters(subidx{:}) = scatter(xScatter,yScatter,obj.scatterMarker{gidx{:}}); set(obj.handles.scatters(subidx{:}),'Parent',obj.handles.axes); end end function drawMean(obj,subidx) %DRAWMEAN drawn the means try % to delete the handles first delete(obj.handles.means(subidx{:})); catch end if obj.showMean % x position X = obj.xticks(subidx{2}) + obj.getOffset(subidx); % plot obj.handles.means(subidx{:}) = plot(X,obj.statistics.mean(subidx{:}),'Parent',obj.handles.axes); end end function drawAddPrctiles(obj,subidx) %DRAWADDPRCTILES draw additional percentiles try delete(obj.handles.addPrctiles(subidx{:})) catch end if ~isempty(obj.addPrctiles) subidx2 = subidx; % x position X = obj.xticks(subidx{2}) + obj.getOffset(subidx); % plot h = zeros(length(obj.addPrctiles),1); for m = 1:length(obj.addPrctiles) subidx2{1} = m; Y = obj.statistics.addPrctiles(subidx2{:}); h(m) = plot(X,Y); text(X,Y,obj.addPrctilesLabels{m},... 'horizontalAlignment','left',... 'verticalAlignment','bottom') end subidxAll = subidx; subidxAll{1} = ':'; obj.handles.addPrctiles(subidxAll{:}) = h; end end function drawGroupGraphics(obj,subidx,gidx) %DRAWGROUPGRAPHICS set graphics options for each group % violin colors if obj.showViolin set(obj.handles.violin(subidx{:}),... 'FaceColor',obj.violinColor{gidx{:}},... 'EdgeColor',obj.lineColor{gidx{:}}); end % box colors set(obj.handles.box(subidx{:}),... 'FaceColor',obj.boxColor{gidx{:}},... 'EdgeColor',obj.lineColor{gidx{:}}); % notch line if obj.notchLine set(obj.handles.notchLowerLines(subidx{:}),... 'linestyle',obj.notchLineStyle{gidx{:}},'color',obj.notchLineColor{gidx{:}}); set(obj.handles.notchUpperLines(subidx{:}),... 'linestyle',obj.notchLineStyle{gidx{:}},'color',obj.notchLineColor{gidx{:}}); end % median line set(obj.handles.medianLines(subidx{:}),'color',obj.medianColor{gidx{:}}); % whiskers set([obj.handles.lowerWhiskers(subidx{:}) obj.handles.upperWhiskers(subidx{:})],... 'color',obj.lineColor{gidx{:}}); set([obj.handles.lowerWhiskers(subidx{:}) obj.handles.upperWhiskers(subidx{:})],... 'linestyle',obj.lineStyle{gidx{:}}); if ~obj.showViolin set([obj.handles.lowerWhiskerTips(subidx{:}) obj.handles.upperWhiskerTips(subidx{:})],... 'color',obj.lineColor{gidx{:}}); end % outliers if ~isempty(obj.statistics.outliers{subidx{:}}) && obj.showOutliers % don't bother if no data set(obj.handles.outliers(subidx{:}),... 'marker',obj.symbolMarker{gidx{:}},'MarkerEdgeColor',obj.symbolColor{gidx{:}}); end % scatter if obj.showScatter set(obj.handles.scatters(subidx{:}),'MarkerEdgeColor',obj.scatterColor{gidx{:}}); end % means if obj.showMean set(obj.handles.means(subidx{:}),... 'color',obj.meanColor{gidx{:}},... 'marker',obj.meanMarker{gidx{:}}); end % additional percentiles if ~isempty(obj.addPrctiles) subidx2 = subidx; for m = 1:length(obj.addPrctiles) subidx2{1} = m; set(obj.handles.addPrctiles(subidx2{:}),... 'color',obj.addPrctilesColors{m},... 'marker',obj.addPrctilesMarkers{m}); end end end function drawGlobalGraphics(obj) %DRAWGLOBALGRAPHICS set global graphics options % violin if obj.showViolin set(obj.handles.violin, ... 'LineWidth',obj.lineWidth, ... 'FaceAlpha',obj.violinAlpha); end % box colors set(obj.handles.box,... 'FaceAlpha',obj.boxAlpha,... 'LineWidth',obj.lineWidth); % sample size if obj.sampleSize set(obj.handles.samplesTxt,'fontsize',obj.sampleFontSize); end % notch line if obj.notchLine set(obj.handles.notchLowerLines,'linewidth',obj.lineWidth); set(obj.handles.notchUpperLines,'linewidth',obj.lineWidth); end % median line set(obj.handles.medianLines,'linewidth',obj.lineWidth); % whiskers set([obj.handles.lowerWhiskers obj.handles.upperWhiskers],... 'linewidth',obj.lineWidth); if ~obj.showViolin set([obj.handles.lowerWhiskerTips obj.handles.upperWhiskerTips],... 'linewidth',obj.lineWidth); end % outliers if obj.showOutliers && isfield(obj.handles,'outliers') set(findobj(obj.handles.outliers,'Type','Scatter'),... 'SizeData',obj.outlierSize); end % scatter if obj.showScatter && isfield(obj.handles,'scatters') set(obj.handles.scatters,'SizeData',obj.scatterSize); if obj.scatterAlpha<1 try % older versions do not support MarkerEdgeAlpha set(obj.handles.scatters,'MarkerEdgeAlpha',obj.scatterAlpha); catch warning('This version of Matlab does not support scatter marker transparency.') end end end % means if obj.showMean && isfield(obj.handles,'means') set(obj.handles.means,'markersize',obj.meanSize); end % sample size if obj.sampleSize && isfield(obj.handles,'groupsTxt') set(obj.handles.groupsTxt,'FontSize',obj.groupLabelFontSize); end if ~isempty(obj.addPrctiles) && isfield(obj.handles,'addPrctilesTxt') set(obj.handles.addPrctilesTxt,'FontSize',obj.addPrctilesTxtSize); end end function drawXsepatator(obj) %DRAWXSEPARATOR draw the x separator lines try % to delete the handles first delete(obj.handles.xseps); catch end if obj.xSeparator xlines = obj.xticks(1:end-1)+0.5.*diff(obj.xticks); % line positions if(obj.rescaleGroups) for it = 1:(length(obj.groupSizes)-1) subidxPre = {1 it obj.groupSizes(it)}; subidxPost = {1 it+1 1}; xlines(it) = (obj.xticks(subidxPre{2}) + obj.getOffset(subidxPre))+(obj.xticks(subidxPost{2}) + obj.getOffset(subidxPost) - (obj.xticks(subidxPre{2}) + obj.getOffset(subidxPre)))/2; end end obj.handles.xseps = zeros(size(xlines)); % handles for n = 1:length(xlines) % draw lines obj.handles.xseps(n) = line([xlines(n) xlines(n)],get(obj.handles.axes,'ylim'),'color',.5+zeros(1,3)); end end end function drawStyle(obj) %DRAWSTYLE draw the style options try % to delete the handles first delete(obj.handles.groupsTxt); catch end % get outermost box widths (correct xlim for this) subidxFirst = cell(1,length(obj.outDims)); [subidxFirst{:}] = ind2sub(obj.outDims, 1); [~,halfboxwidthFirst] = calcBoxWidths(obj,subidxFirst); subidxLast = cell(1,length(obj.outDims)); [subidxLast{:}] = ind2sub(obj.outDims, prod(obj.outDims)); [~,halfboxwidthLast] = calcBoxWidths(obj,subidxLast); % set default x axis set(obj.handles.axes,'xtick',obj.xticks,... 'xlim',[min(obj.xticks)-0.5*obj.diffx-halfboxwidthFirst max(obj.xticks)+0.5*obj.diffx+halfboxwidthLast],... 'xticklabel',obj.xlabels); if(obj.rescaleGroups) tmpTicks = zeros(size(obj.groupSizes)); for it = 1:(length(obj.groupSizes)) subidxPre = {1 it 1}; subidxPost = {1 it obj.groupSizes(it)}; tmpTicks(it) = (obj.xticks(subidxPre{2}) + obj.getOffset(subidxPre))+(obj.xticks(subidxPost{2}) + obj.getOffset(subidxPost) - (obj.xticks(subidxPre{2}) + obj.getOffset(subidxPre)))/2; end set(obj.handles.axes,'xtick',tmpTicks); end switch lower(obj.style) case 'normal' % do nothing case 'hierarchy' % check grouplabels options if ~isempty(obj.groupLabels) assert(isvector(obj.groupLabels) && iscell(obj.groupLabels),... 'iosr:boxPlot:groupLabelsType', ... 'The GROUPLABELS option should be a cell vector'); if ~isempty(obj.outDims(3:end)) assert(isequal(obj.outDims(3:end),cellfun(@length,obj.groupLabels)),... 'iosr:boxPlot:groupLabelsSize', ... ['The GROUPLABELS option should be a cell vector; ' ... 'the Nth element should contain a vector of length SIZE(Y,N+2)']) else assert(cellfun(@length,obj.groupLabels)==1,... 'iosr:boxPlot:groupLabelsSize', ... 'The GROUPLABELS option should have length 1 if no grouping dimensions are specified.') end end % number of label rows labelrows = length(obj.outDims)-1; ylim = get(obj.handles.axes,'ylim'); % need y limitis if isnumeric(obj.groupLabelHeight) labelheight = obj.groupLabelHeight; else % determine label height labelheight = obj.calcLabelHeight(); % y range in units calculated based on normalised height end % determine y centre points of labels based on how many labels % and their height ymids = ylim(1) - (labelrows:-1:1).*labelheight + 0.5.*labelheight; % clear axis tick labels set(obj.handles.axes,'xticklabels',[]); % put labels on plot totalticks = prod(obj.outDims(2:end)); % total number of ticks across allticklocs = sort(obj.groupXticks(:)); % their locations obj.handles.groupsTxt = []; for c = labelrows:-1:1 % work down hierarchy Y = ymids(c); % y location nskip = totalticks/prod(obj.outDims(2:c+1)); % skip through tick locations based on hierarchy first = 1:nskip:totalticks; % first tick in group last = nskip:nskip:totalticks; % last tick in group labellocs = allticklocs(first) + 0.5.*(allticklocs(last)-allticklocs(first)); % centre point for label location % work through each label location for n = 1:length(labellocs) if c==1 % special case: x axis ticks gLabel = obj.xlabels{n}; else % every other group if isempty(obj.groupLabels) % auto increment label gLabel = num2str(mod(n-1,obj.outDims(c+1))+1); else % get from input gLabel = obj.groupLabels{c-1}(mod(n-1,obj.outDims(c+1))+1); % cyclic index into labels if isnumeric(gLabel) % convert numbers to strings gLabel = num2str(gLabel); end end end % place actual label t = text(labellocs(n),Y,gLabel,'HorizontalAlignment','center','VerticalAlignment','middle'); % store info about placement setappdata(t,'rows',labelrows); setappdata(t,'row',c); % add to handles obj.handles.groupsTxt = [obj.handles.groupsTxt t]; end end otherwise error('iosr:boxPlot:unknownStyle','Unkown ''style'' option.') end end function drawLegend(obj) %DRAWLEGEND draw the legend try % to delete the handles first delete(obj.handles.legend); catch end if obj.showLegend % dimensions of group data gDims = cellfun(@length,obj.groupLabels); if isempty(gDims) gDims = 1; end subidx = cell(1,length(gDims)); % going to change order of handles orderidx = cell(1,length(gDims)+1); order = zeros(prod(gDims),1); % preallocate legend labels lgndstr = cell(prod(gDims),1); for n = 1:prod(gDims) % use linear index to get group data [subidx{:}] = ind2sub([gDims 1], n); % order looks at trailing dimensions first [orderidx{:}] = ind2sub(fliplr([gDims 1]), n); orderidx = fliplr(orderidx); order(n) = sub2ind([gDims 1],orderidx{:}); % create labels start = true; % create start of string label for m = length(subidx):-1:1 % work from end through dimensions str = obj.ensureString(obj.groupLabels{m}(subidx{m})); if start % initialise label lgndstr{n} = str; start = false; else % prepend string label lgndstr{n} = [str ', ' lgndstr{n}]; end end end % choose target for the legend if ~obj.arrayElementsEqual(obj.boxColor) legendTarget = obj.handles.box; target = 'box'; elseif ~obj.arrayElementsEqual(obj.medianColor) legendTarget = obj.handles.medianLines; target = 'medianLines'; elseif obj.showViolin && ~obj.arrayElementsEqual(obj.violinColor) legendTarget = obj.handles.violin; target = 'violin'; else error('iosr:boxPlot:legend','The legend uses the box, median line colors, or violins to create legend entries. However, these items appear to be identical, which would make the legend impossible to interpret. Change the ''boxColor'', ''medianColor'', or ''violinColor'' option.'); end % create legend in specified order if ~isempty(legendTarget) [obj.handles.legend, icons] = legend(legendTarget(1,1,order),lgndstr(order)); set(obj.handles.legend,'location','best'); % add alpha to legend entries PatchInLegend = findobj([obj.handles.legend; icons(:)], 'type', 'patch'); if ~isempty(PatchInLegend) switch lower(target) case 'box' set(PatchInLegend, 'facealpha', obj.boxAlpha); case 'violin' set(PatchInLegend, 'facealpha', obj.violinAlpha); end end end end end function createTheme(obj) %CREATETHEME create the color/line theme switch lower(obj.theme) % choose scheme case 'colorall' boxalpha = 0.4; boxcolor = obj.themeColors; linecolor = 'k'; linestyle = '-'; meancolor = 'k'; mediancolor = 'k'; notchlinecolor = obj.themeColors; notchlinestyle = ':'; scattercolor = obj.themeColors; symbolcolor = obj.themeColors; case {'colorlines', 'graylines'} if strcmpi(obj.theme,'graylines') warning('iosr:boxPlot:graylines', 'The ''graylines'' theme is deprecated. Use the ''colorlines'' theme instead, and specify ''themeColors'' as @gray.') end boxalpha = obj.boxAlpha; boxcolor = 'none'; linecolor = 'k'; linestyle = '-'; meancolor = 'k'; mediancolor = obj.themeColors; notchlinecolor = obj.themeColors; notchlinestyle = ':'; scattercolor = obj.themeColors; symbolcolor = obj.themeColors; case {'colorboxes', 'grayboxes'} if strcmpi(obj.theme,'grayboxes') warning('iosr:boxPlot:grayboxes', 'The ''grayboxes'' theme is deprecated. Use the ''colorboxes'' theme instead, and specify ''themeColors'' as @gray.') end boxalpha = obj.boxAlpha; boxcolor = obj.themeColors; linecolor = 'k'; linestyle = '-'; meancolor = 'k'; mediancolor = 'k'; notchlinecolor = 'k'; notchlinestyle = ':'; scattercolor = [.5 .5 .5]; symbolcolor = 'k'; otherwise if ~strcmpi(obj.theme,'default') warning(['Unknown theme ''' obj.theme ''' specified. Using default']) end boxalpha = 1; boxcolor = 'none'; linecolor = 'k'; linestyle = '-'; meancolor = obj.themeColors; mediancolor = obj.themeColors; notchlinecolor = 'k'; notchlinestyle = ':'; scattercolor = [.5 .5 .5]; symbolcolor = obj.themeColors; end % turn off legend while changing parameters legendState = obj.showLegend; obj.showLegend = false; % set parameters obj.boxAlpha = boxalpha; obj.boxColor = boxcolor; obj.medianColor = mediancolor; obj.lineColor = linecolor; obj.lineStyle = linestyle; obj.meanColor = meancolor; obj.notchLineColor = notchlinecolor; obj.notchLineStyle = notchlinestyle; obj.scatterColor = scattercolor; obj.symbolColor = symbolcolor; % turn legend back on if it was on obj.showLegend = legendState; end function calculateStats(obj) %CALCULATESTATS calculate the statistic for the box plot calculateStats@iosr.statistics.statsPlot(obj); % calculate stats obj.statistics.percentile = obj.percentile; % percentile outsize = size(obj.statistics.median); obj.statistics.addPrctiles = zeros([length(obj.addPrctiles) outsize(2:end)]); obj.statistics.PL = zeros(outsize); % lower percentile obj.statistics.PU = zeros(outsize); % upper percentile subidx = cell(1,length(obj.outDims)); for n = 1:prod(obj.outDims) [subidx{:}] = ind2sub(obj.outDims, n); subidxAll = subidx; subidxAll{1} = ':'; if ~isempty(obj.addPrctiles) obj.statistics.addPrctiles(subidxAll{:}) = iosr.statistics.quantile(obj.y(subidxAll{:}),obj.addPrctiles(:)./100,[],obj.method,obj.weights(subidxAll{:})); end obj.statistics.PL(subidx{:}) = iosr.statistics.quantile(obj.y(subidxAll{:}),min(obj.percentile)/100,[],obj.method,obj.weights(subidxAll{:})); obj.statistics.PU(subidx{:}) = iosr.statistics.quantile(obj.y(subidxAll{:}),max(obj.percentile)/100,[],obj.method,obj.weights(subidxAll{:})); end % check for notches extending beyond box if (any(obj.statistics.notch_u(:)>obj.statistics.Q3(:)) || any(obj.statistics.notch_l(:)0.33,1,'first'); if numel(out)<=N cmap = cmapTest(round(linspace(firstLumaIX,N,numel(out))),:); % sample color map else pad = ceil((firstLumaIX/N)*numel(out)); % pad for the black region cmap = color(numel(out)+pad); % colormap cmap = cmap(pad+1:end,:); % remove padded region end case 'line' % ensure maximum luminance of 0.66 lastLumaIX = find(YL<0.66,1,'last'); if numel(out)<=N cmap = cmapTest(round(linspace(1,lastLumaIX,numel(out))),:); % sample color map else pad = ceil(((N-lastLumaIX)/N)*numel(out)); % pad for the black region cmap = color(numel(out)+pad); % colormap cmap = cmap(1:end-pad,:); % remove padded region end end else cmap = color(numel(out)); % colormap end for n = 1:numel(out) out{n} = cmap(n,:); end end end function [boxwidth, halfboxwidth] = calcBoxWidths(obj,subidx) %CALCBOXWIDTHS calculate the box width % size of boxes if ischar(obj.boxWidth) boxwidth = 0.75*(obj.groupRange/prod(obj.groupDims)); else boxwidth = obj.boxWidth; end if obj.scaleWidth % scale box width according to sample size maxN = max(obj.statistics.N(:)); halfboxwidth = 0.5 * (sqrt(obj.statistics.N(subidx{:})/maxN)*boxwidth); else % normal box width halfboxwidth = boxwidth/2; end end function halfviolinwidth = calcHalfViolinWidth(obj,subidx) %HALFVIOLINWIDTH calculate the violin width % size of boxes if ischar(obj.violinWidth) halfviolinwidth = 0.75*(obj.groupRange/prod(obj.groupDims)); else halfviolinwidth = obj.violinWidth; end if obj.scaleWidth % scale box width according to sample size maxN = max(obj.statistics.N(:)); halfviolinwidth = 0.5 * (sqrt(obj.statistics.N(subidx{:})/maxN)*halfviolinwidth); else halfviolinwidth = halfviolinwidth / 2; end end function labelheight = calcLabelHeight(obj) %CALCLABELHEIGHT calculate the label height for the axes ylim = get(obj.handles.axes,'ylim'); plotheight = get(obj.handles.axes,'position'); % plot position plotheight = plotheight(4); % plot height in normalised units plotrange = abs(diff(ylim)); % y range labelheight = plotheight*plotrange*0.05; % y range in units calculated based on normalised height end function offset = getOffset(obj,subidx) %GET_OFFSET calculate offset for hierarchical data dims = obj.outDims(3:end); subidx = subidx(3:end); if isempty(dims) || prod(dims)==1 offset = 0; else boxspacing = obj.groupRange/prod(obj.groupDims); offsetn = (0:prod(dims)-1)-((prod(dims)-1)./2); n = sub2ind([dims(end:-1:1) 1],subidx{end:-1:1}); offset = offsetn(n).*boxspacing; end end function val = checkAddPrctileProp(obj,init,default) %CHECKADDPRCTILEPROP check and set default values of additional percentile properties val = init(:); currlength = length(val); if currlength=0) && all(cOutput(:)<=1), 'iosr:boxPlot:colorFhandleRange', ['The color ''' char(fHandle) ''' function does not appear to return RGB values in the interval [0,1].']) end end function option = replaceAuto(option, default, errorMsg) %REPLACEAUTO check for and replace auto parameters if ischar(option) if strcmpi(option, 'auto') option = default; else error('iosr:boxPlot:invalidOption',errorMsg) end end end end end