DXDENSITY procedure

Produces one-dimensional density (or violin) plots (D. B. Baird).

Options

BANDWIDTH = scalar	Bandwidth for kernel smoothing (0-1); default density is chosen according to the number of observations
GAP = scalar	The size of the gap (0-1) between envelopes when there are several densities; default 0.1
TRANSFORM = string token	Transformed scale for the data (identity, log, log10, logit, probit, cloglog, square, exp, exp10, ilogit, iprobit, icloglog, root); default is to use the transform defined for XAXIS
AXISTITLE = text	The title for the data axis; default is the name of the DATA variate
GROUPSTITLE = text	The title for the groups or variates axis; default is to use the name of the GROUPS factor
WINDOW = scalar	Window number for the graph; default 3
ORIENTATION = string token	Orientation of plots (horizontal, vertical); default vert
METHOD = string token	Method for plotting the density envelope (fill, line); default fill
SCREEN = string token	Whether to clear screen before the plot (clear, keep, resize); default clea

Parameters

DATA = variates or pointers	The data whose density is to be plotted
GROUPS = factors	Factor to divide values of a single variate into groups; default * i.e. none
TITLE = texts	Title for graph; default uses the names of the data variates and type of plot

Description

DXDENSITY produces density (or violin) plots, using high-resolution graphics. The data are specified by the DATA parameter. For a single density plot, DATA should be set to a variate. To plot several densities, you can set DATA to a pointer of variates. Alternatively, it can be set to a single variate, with the GROUPS parameter then specifying a factor to identify groups of points whose densities are to be plotted separately.

The points are plotted along a line, with a kernel density smooth on either side to indicate the density of points along the line. The BANDWIDTH option specifies the band width for the kernel smoothing; larger values make it smoother, and smaller values allow it to be rougher. The default is chosen automatically, according to the number of observations. The gap between the envelopes for different variates or groups can be defined by the GAP option; this must be between 0 and 1 (default 0.1).

The METHOD option controls how the density envelope is drawn around the points, either as a filled region or as a line. You can use the TITLE parameter to supply an overall title for the plot. The AXISTITLE and GROUPSTITLE options can be used to supply titles for the data and groups axes respectively. The WINDOW option specifies the number of the window to use for the plot (default 3), and the SCREEN option controls whether the screen is cleared first, as usual (see DGRAPH).

The data can be transformed by using the TRANSFORM option. If this is not set, DXDENSITY uses the setting of TRANSFORM, defined by the XAXIS directive for the specified WINDOW. The available settings are the same as those of the TRANSFORM option of XAXIS.

The ORIENTATION option controls whether the data axis is horizontal or vertical (default). The XAXIS and YAXIS directives can be used, prior to using DXDENSITY, to set attributes of the axes of the plot in the window that is to be used. Note that, when the orientation is horizontal, the attributes of the x- and y-axes are swapped, so that the x-axis settings are always applied to the data axis.

Options: BANDWIDTH, GAP, TRANSFORM, AXISTITLE, GROUPSTITLE, WINDOW, ORIENTATION, METHOD, SCREEN.

Parameters: DATA, GROUPS, TITLE.

Action with RESTRICT

If any of the variates or factors are restricted, only the units not excluded by the restriction will be plotted.

See also

Directive: DHISTOGRAM.

Procedures: DXYDENSITY, DOTHISTOGRAM, BOXPLOT, KERNELDENSITY, RUGPLOT.

Commands for: Graphics.

Example

CAPTION   'DXDENSITY Example',\
          !t('1) The three variates America, Asia_Oc & Other contain the',\
          'heights of volcanoes in three regions of the world.');\ 
          STYLE=meta,plain
VARIATE   [VALUES=199,197,193,185,177,172,157,156,140,140,130,126,124,124,113,\
          102,100,102,94,93,89,86,83,83,83,82,77,73,70,62,58,51,51,42,40,\
          34,36,67,67,66,60,57,57,53,49,43,43,40,35,35] America
&         [VALUES=156,137,125,122,120,112,109,103,100,100,96,95,95,90,83,\
          81,81,81,77,75,75,73,71,71,67,66,66,64,62,60,60,60,59,58,57,\
          56,56,55,54,54,52,52,52,51,50,49,49,48,45,44,44,37,36,36,26,\
          26,24,19,11,10,41] Asia_Oc
&         [VALUES=134,125,114,111,100,90,80,75,49,21,21,30,60,17,19] Other
DXDENSITY !P(America,Asia_Oc,Other)
CAPTION   !T('2) The three sets of heights are combined into a single',\
          'variate and a factor is set up to specify which values in the',\
          'combined variate came from which region.')
VARIATE   [VALUES=#America,#Asia_Oc,#Other] All
FACTOR    [LABELS=!t(America,'Asia/Oceania',Elsewhere);\ 
          VALUES=50(1),61(2),15(3)] Region
DXDENSITY [BANDWIDTH=0.35; AXISTITLE='Height of volcano'; GROUPSTITLE='Region']\
          All; GROUP=Region; TITLE='Volcano height by region'