Categories &

Functions List

Function Reference: parallelplot

statistics: parallelplot (data)

statistics: parallelplot (tbl)

statistics: parallelplot (parent, …)

statistics: parallelplot (…, Name, Value)

statistics: h = parallelplot (…)

Parallel coordinates plot of observations of several variables.

parallelplot (data) draws each column of the numeric matrix data as a vertical ruler, a coordinate, and each row as a line joining its values across the rulers. By default each ruler has limits of its own, from the smallest to the largest value of its column.

parallelplot (tbl) draws the variables of the table tbl as coordinates, named after them. A variable of categories, text or logical values places its distinct values evenly along its ruler.

'CoordinateData' chooses the columns of data to draw, and 'CoordinateVariables' the variables of tbl, by names, indices or a logical vector. With 'GroupData', a vector with one element for each observation, or 'GroupVariable', a variable of tbl, each group is drawn in a colour of its own, with a legend naming the groups in the order in which each first appears. 'DataNormalization' places the values as they are on rulers of their own, 'range', or on one shared scale: as they are, 'none'; as z-scores, 'zscore'; divided by the standard deviation, 'scale'; less the mean, 'center'; divided by the 2-norm, 'norm'.

parallelplot (parent, …) places the chart in the figure or panel parent rather than the current figure. The chart takes the place of the current axes of its parent, and is refused where hold is on for them.

parallelplot (…, Name, Value) sets properties of the chart; see stats.chart.ParallelCoordinatesPlot for all of them.

h = parallelplot (…) returns the chart, a stats.chart.ParallelCoordinatesPlot object.

See also: stats.chart.ParallelCoordinatesPlot, parallelcoords

Source Code: parallelplot

The four measurements of each iris flower, joined across four rulers.

 load fisheriris
 parallelplot (meas);
plotted figure

Grouped by species, with the columns named, as z-scores on one scale.

 load fisheriris
 parallelplot (meas, 'GroupData', species, 'CoordinateTickLabels', ...
               {'Sepal length', 'Sepal width', 'Petal length', ...
                'Petal width'}, 'DataNormalization', 'zscore');
plotted figure

From a table, with a variable of categories as a coordinate.

 load carsmall
 t = table (MPG, Horsepower, Weight, categorical (cellstr (Origin)), ...
            'VariableNames', {'MPG', 'Horsepower', 'Weight', 'Origin'});
 parallelplot (t, 'GroupVariable', 'Origin', 'CoordinateVariables', ...
               {'MPG', 'Horsepower', 'Weight'});
plotted figure