Stata 11 help for biplot

help biplot dialog: biplot -------------------------------------------------------------------------------

Title

[MV] biplot -- Biplots

Syntax

biplot varlist [if] [in] [, options]

options description ------------------------------------------------------------------------- Main rowover(varlist) identify observations from different groups of varlist; may not be combined with separate or norow dim(# #) two dimensions to be displayed; default dim(2 1) std use standardized instead of centered variables alpha(#) row weight=#; column weight=1-#; default is 0.5 stretch(#) stretch the column (variable) arrows mahalanobis approximate Mahalanobis distance; implies alpha(0) xnegate negate the data relative to the x axis ynegate negate the data relative to the y axis autoaspect adjust aspect ratio on the basis of the data; default aspect ratio is 1 separate produce separate plots for rows and columns; may not be combined with rowover() nograph suppress graph table display table showing biplot coordinates

Rows rowopts(row_options) affect rendition of rows (observations) row#opts(row_options) affect rendition of rows (observations) in the #th group of varlist defined in rowover(); available only with rowover() rowlabel(varname) specify label variable for rows (observations) norow suppress row points; may not be combined with rowover() generate(newvar_x newvar_y) store biplot coordinates for observations in variables newvar_x and newvar_y

Columns colopts(col_options) affect rendition of columns (variables) negcol include negative column (variable) arrows negcolopts(col_options) affect rendition of negative columns (variables) nocolumn suppress column arrows

Y axis, X axis, Titles, Legend, Overall twoway_options any options other than by() documented in [G] twoway_options -------------------------------------------------------------------------

row_options description ------------------------------------------------------------------------- marker_options change look of markers (color, size, etc.) marker_label_options change look or position of marker labels nolabel remove the default row (variable) label from the graph name(name) override the default name given to rows (observations) -------------------------------------------------------------------------

col_options description ------------------------------------------------------------------------- pcarrow_options affect the rendition of paired-coordinate arrows nolabel remove the default column (variable) label from the graph name(name) override the default name given to columns (variables) -------------------------------------------------------------------------

Menu

Statistics > Multivariate analysis > Biplot

Description

biplot displays a two-dimensional biplot of a dataset. A biplot simultaneously displays the observations (rows) and the relative positions of the variables (columns). Marker symbols (points) are displayed for observations and arrows are displayed for variables. Observations are projected to two dimensions such that the distance between the observations is approximately preserved. The cosine of the angle between arrows approximates the correlation between the variables.

Options

+------+ ----+ Main +-------------------------------------------------------------

rowover(varlist) distinguishes groups among observations (rows) by highlighting observations on the plot for each group identified by equal values of the variables in varlist. By default, the graph contains a legend that consists of group names. rowover() may not be combined with separate or norow.

dim(# #) identifies the dimensions to be displayed. For instance, dim(3 2) plots the third dimension (vertically) versus the second dimension (horizontally). The dimension numbers cannot exceed the number of variables. The default is dim(2 1).

std produces a biplot of the standardized variables instead of the centered variables.

alpha(#) specifies that the variables be scaled by lambda^# and the observations by lambda^(1-#), where lambda are the singular values. It is required that 0 <= # <= 1. The most common values are 0, 0.5, and 1. The default is alpha(0.5) and is known as the symmetrically scaled biplot or symmetric factorization biplot. The result with alpha(1) is the principal-component biplot, also called the row-preserving metric (RPM) biplot. The biplot with alpha(0) is referred to as the column-preserving metric (CPM) biplot.

stretch(#) causes the length of the arrows to be multiplied by #. For example, stretch(1) would leave the arrows the same length, stretch(2) would double their length, and stretch(0.5) would halve their length.

mahalanobis implies alpha(0) and scales the positioning of points (observations) by sqrt(n-1) and positioning of arrows (variables) by 1/sqrt(n-1). This additional scaling causes the distances between observations to change from being approximately proportional to the Mahalanobis distance to instead being approximately equal to the Mahalanobis distance. Also, the inner products between variables approximate their covariance.

xnegate specifies that dimension-1 (x axis) values be negated (multiplied by -1).

ynegate specifies that dimension-2 (y axis) values be negated (multiplied by -1).

autoaspect specifies that the aspect ratio be automatically adjusted based on the range of the data to be plotted. This option can make some biplots more readable. By default, biplot uses an aspect ratio of one, producing a square plot. Some biplots will have little variation in the y-axis direction, and using the autoaspect option will better fill the available graph space while preserving the equivalence of distance in the x and y axes.

As an alternative to autoaspect, the twoway_option aspectratio() can be used to override the default aspect ratio. biplot accepts the aspectratio() option as a suggestion only and will override it when necessary to produce plots with balanced axes; i.e., distance on the x axis equals distance on the y axis.

twoway_options, such as xlabel(), xscale(), ylabel(), and yscale(), should be used with caution. These options are accepted but may have unintended side effects on the aspect ratio.

separate produces separate plots for the row and column categories. The default is to overlay the plots. separate may not be combined with rowover().

nograph suppresses displaying the graph.

table displays a table with the biplot coordinates.

+------+ ----+ Rows +-------------------------------------------------------------

rowopts(row_options) affects the rendition of the points plotting the rows (observations). This option may not be combined with rowover(). The following row_options are allowed:

marker_options affect the rendition of markers drawn at the plotted points, including their shape, size, color, and outline; see [G] marker_options.

marker_label_options specify the properties of marker labels; see [G] marker_label_options. mlabel() in rowopts() may not be combined with the rowlabel() option.

nolabel removes the default row label from the graph.

name(name) overrides the default name given to rows.

row#opts(row_options) affects rendition of the points plotting the rows (observations) in the #th group identified by equal values of the variables in varlist defined in rowover(). This option requires specifying rowover(). See rowopts() above for the allowed row_options, except mlabel() is not allowed with row#opts().

rowlabel(varname) specifies label variable for rows (observations).

norow suppresses plotting of row points. This option may not be combined with rowover().

generate(newvar_x newvar_y) stores biplot coordinates for rows in variables newvar_x and newvar_y.

+---------+ ----+ Columns +----------------------------------------------------------

colopts(col_options) affects the rendition of the arrows and points plotting the columns (variables). The following col_options are allowed:

pcarrow_options affect the rendition of paired-coordinate arrows; see > [G] graph twoway pcarrow.

nolabel removes the default column label from the graph.

name(name) overrides the default name given to columns.

negcol includes negative column (variable) arrows on the plot.

negcolopts(col_options) affects the rendition of the arrows and points plotting the negative columns (variables). The col_options allowed are given above.

nocolumn suppresses plotting of column arrows.

+-----------------------------------------+ ----+ Y axis, X axis, Titles, Legend, Overall +--------------------------

twoway_options are any of the options documented in [G] twoway_options excluding by(). These include options for titling the graph (see [G] title_options) and for saving the graph to disk (see [G] saving_option). See autoaspect above for a warning against using options such as xlabel(), xscale(), ylabel(), and yscale().

Examples

Setup . webuse renpainters

Biplot of 10 Renaissance painters on four attributes . biplot composition-expression, alpha(1) stretch(10) table rowlabel(painter) rowopts(name(Painters)) colopts(name(Attributes)) title(Renaissance painters) autoaspect

Same as above, but highlights observations with the same score on attribute colour . biplot composition-expression, alpha(1) stretch(10) table rowover(colour) rowlabel(painter) colopts(name(Attributes)) title(Renaissance painters) autoaspect

Same as above, but uses the defined names and suppresses the marker label . biplot composition-expression, alpha(1) stretch(10) table rowover(colour) row1opts(name("Color score: 4") nolabel) row2opts(name("Color score: 7") nolabel) row3opts(name("Color score: 8") nolabel) row4opts(name("Color score: 9") nolabel) row5opts(name("Color score: 10") nolabel) row6opts(name("Color score: 12") nolabel) row7opts(name("Color score: 16") nolabel) colopts(name(Attributes)) title(Renaissance painters) autoaspect

Saved results

biplot saves the following in r():

Scalars r(rho1) explained variance by component 1 r(rho2) explained variance by component 2 r(rho) total explained variance r(alpha) value of alpha() option

Matrices r(U) biplot coordinates for the observations; stored only if the row dimension does not exceed Stata's maximum matrix size; as an alternative, use generate() to store biplot coordinates for the observations in variables r(V) biplot coordinates for the variables r(Vstretch) biplot coordinates for the variables times stretch() factor

Also see

Manual: [MV] biplot

Help: [MV] ca, [MV] mds, [MV] pca


© Copyright 1996–2009 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index