help mca postestimation dialogs: predict estat
mcaplot mcaprojection
screeplot
also see: mca
-------------------------------------------------------------------------------
Title
[MV] mca postestimation -- Postestimation tools for mca
Description
The following postestimation commands are of special interest after mca:
command description
-------------------------------------------------------------------------
mcaplot plot of category coordinates
mcaprojection MCA dimension projection plot
estat coordinates display of category coordinates
estat subinertia matrix of inertias of the active variables (after
JCA only)
estat summarize estimation sample summary
screeplot plot principal inertias (eigenvalues)
-------------------------------------------------------------------------
The following standard postestimation commands are also available:
command description
-------------------------------------------------------------------------
* estimates cataloging estimation results
predict row and category coordinates
-------------------------------------------------------------------------
* All estimates subcommands except table and stats are available.
Special-interest postestimation commands
mcaplot produces a scatterplot of category points of the MCA variables in
two dimensions.
mcaprojection produces a projection plot of the coordinates of the
categories of the MCA variables.
estat coordinates displays the category coordinates, optionally with
column statistics.
estat subinertia displays the matrix of inertias of the active variables
(after JCA only).
estat summarize displays summary information of MCA variables over the
estimation sample.
Syntax for predict
predict [type] newvar [if] [in] [, statistic normalize(norm)
dimensions(#)]
predict [type] {stub*|newvarlist} [if] [in] [, statistic
normalize(norm) dimensions(numlist)]
statistic description
-------------------------------------------------------------------------
Main
rowscores row scores (coordinates), the default
score(varname) scores (coordinates) for MCA variable varname
-------------------------------------------------------------------------
norm description
-------------------------------------------------------------------------
standard use standard normalization
prinicpal use principal normalization
-------------------------------------------------------------------------
Menu
Statistics > Postestimation > Predictions, residuals, etc.
Options for predict
+------+
----+ Main +-------------------------------------------------------------
rowscores specifies that row scores (row coordinates) be computed. The
row scores returned are based on the indicator matrix approach to
multiple correspondence analysis, even if another method was
specified in the original mca estimation. The sample for which row
scores are computed may exceed the estimation sample; e.g., it may
include supplementary rows (variables). score() and rowscores are
mutually exclusive. rowscores is the default.
score(varname) specifies the name of a variable from the preceding MCA
for which scores should be computed. The variable may be a regular
categorical variable, a crossed variable, or a supplementary
variable. score() and rowscores are mutually exclusive.
+---------+
----+ Options +----------------------------------------------------------
normalize(norm) specifies the normalization of the scores (coordinates).
normalize(standard) returns coordinates in standard normalization.
normalize(principal) returns principal scores. The default is the
normalization method specified with mca during estimation, or
normalize(standard) if no method was specified.
dimensions(#) or dimensions(numlist) specifies the dimensions for which
scores (coordinates) are computed. The number of dimensions
specified should equal the number of variables in newvarlist. If
dimensions() is not specified, scores for dimensions 1,...,k are
returned, where k is the number of variables in newvarlist. The
number of variables in newvarlist should not exceed the number of
dimensions extracted during estimation.
Syntax for estat coordinates
estat coordinates [varlist] [, normalize(norm) stats format(%fmt)]
Note: variables in varlist must be from the preceding mca and may refer
to either a regular categorical variable or a crossed variable. The
variables in varlist may also be chosen from the supplementary variables.
options description
-------------------------------------------------------------------------
normalize(standard) standard coordinates
normalize(principal) principal coordinates
stats include mass, distance, and inertia
format(%fmt) display format; default is format(%9.4f)
-------------------------------------------------------------------------
Menu
Statistics > Postestimation > Predictions, residuals, etc.
Options for estat coordinates
normalize(norm) specifies the normalization of the scores (coordinates).
normalize(standard) returns coordinates in standard normalization.
normalize(principal) returns principal scores. The default is the
normalization method specified with mca during estimation, or
normalize(standard) if no method was specified.
stats includes the column mass, the distance of the columns to the
centroid, and the column inertias in the table.
format(%fmt) specifies the display format for the matrix, e.g.,
format(%8.3f). The default is format(%9.4f).
Syntax for estat subinertia
estat subinertia
Menu
Statistics > Postestimation > Predictions, residuals, etc.
Syntax for estat summarize
estat summarize [, crossed labels noheader noweights ]
options description
-------------------------------------------------------------------------
Main
crossed summarize crossed and uncrossed variables as used
labels display variable labels
noheader suppress the header
noweights ignore weights
-------------------------------------------------------------------------
Menu
Statistics > Postestimation > Predictions, residuals, etc.
Options for estat summarize
+------+
----+ Main +-------------------------------------------------------------
crossed specifies summarizing the crossed variables if crossed variables
are used in the MCA, rather than the crossing variables from which
they are formed. The default is to summarize the crossing variables
and single categorical variables used in the MCA.
labels displays variable labels.
noheader suppresses the header.
noweights ignores the weights, if any. The default when weights are
present is to perform a weighted summarize on all variables except
the weight variable itself. An unweighted summarize is performed on
the weight variable.
Syntax for mcaplot
mcaplot [speclist] [, options]
where
speclist = spec [spec ...]
and
spec = varlist | (varname [, plot_options])
and variables in varlist or varname must be from the preceding mca and
may refer to either a regular categorical variable or a crossed variable.
The variables may also be supplementary.
options description
-------------------------------------------------------------------------
Options
combine_options affect the rendition of the combined graphs
overlay overlay the plots of the variables; default
is to produce separate plots
dimensions(#_1 #_2) display dimensions #_1 and #_2;
default is dimension(2 1)
normalize(standard) display standard coordinates
normalize(principal) display principal coordinates
maxlength(#) use # as maximum number of characters for
labels; default is maxlength(12)
xnegate negate the coordinates relative to the x axis
ynegate negate the coordinates relative to the y axis
origin mark the origin and draw origin axes
originlopts(line_options) affect the rendition of the origin axes
Y axis, X axis, Titles, Legend, Overall
twoway_options any options other than by() documented in
[G] twoway_options
-------------------------------------------------------------------------
plot_options description
-------------------------------------------------------------------------
marker_options change look of markers (color, size, etc.)
marker_label_options add marker labels; change look or position
twoway_options titles, legends, axes, added lines and text,
regions, etc.
-------------------------------------------------------------------------
Menu
Statistics > Multivariate analysis > Correspondence analysis >
Postestimation after MCA or JCA > Plot of category coordinates
Options for mcaplot
+-------+
----+ Plots +------------------------------------------------------------
plot_options affect the rendition of markers, including their shape,
size, color, and outline (see [G] marker_options) and specify if and
how the markers are to be labeled (see [G] marker_label_options).
These options may be specified for each variable. If the overlay
option is not specified, then for each variable you may also specify
many of the twoway_options excluding by(), name(), and aspectratio().
See the remarks with the other twoway_options below warning against
using options such as xlabel(), xscale(), ylabel(), and yscale().
+---------+
----+ Options +----------------------------------------------------------
combine_options affect the rendition of the combined plot; see [G] graph
combine. combine_options may not be specified with overlay.
overlay overlays the biplot graphs for the variables. The default is to
produce a combined graph of the biplot graphs.
dimension(#_1 #_2) identifies the dimensions to be displayed. For
instance, dimension(3 2) plots the third dimension (vertically)
versus the second dimension (horizontally). The dimension number
cannot exceed the number of extracted dimensions. The default is
dimension(2 1).
normalize(norm) specifies the normalization of the coordinates.
normalize(standard) returns coordinates in standard normalization.
normalize(principal) returns principal coordinates. The default is
the normalization method specified with mca during estimation, or
normalize(standard) if no method was specified.
maxlength(#) specifies the maximum number of characters for row and
column labels; the default is maxlength(12).
xnegate specifies that the x-axis coordinates be negated (multiplied by
-1).
ynegate specifies that the y-axis coordinates be negated (multiplied by
-1).
origin marks the origin and draws the origin axes.
originlopts(line_options) affect the rendition of the origin axes. See
[G] line_options.
+-----------------------------------------+
----+ Y axis, X axis, Titles, Legend, Overall +--------------------------
twoway_options are any of the options documented in [G] twoway_options
excluding by(). See the remarks below for a warning against using
options such as xlabel(), xscale(), ylabel(), and yscale().
mcaplot automatically adjusts the aspect ratio on the basis of the
range of the data and ensures that the axes are balanced. As an
alternative, the twoway_option aspectratio() can be used to override
the default aspect ratio. mcaplot 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.
Syntax for mcaprojection
mcaprojection [speclist] [, options]
where
speclist = spec [spec ...]
and
spec = varlist | (varname [, plot_options])
and variables in varlist or varname must be from the preceding mca and
may refer to either a regular categorical variable or a crossed variable.
The variables may also be supplementary.
options description
-------------------------------------------------------------------------
Options
dimensions(numlist) display numlist dimensions; default is all
normalize(principal) scores (coordinates) should be in principal
normalization
normalize(standard) scores (coordinates) should be in standard
normalization
alternate alternate labels
maxlength(#) use # as maximum number of characters for
labels; default is maxlength(12)
combine_options affect the rendition of the combined graphs
Y axis, X axis, Titles, Legend, Overall
twoway_options any options other than by() documented in
[G] twoway_options
-------------------------------------------------------------------------
plot_options description
-------------------------------------------------------------------------
marker_options change look of markers (color, size, etc.)
marker_label_options add marker labels; change look or position
twoway_options titles, legends, axes, added lines and text,
regions, etc.
-------------------------------------------------------------------------
Menu
Statistics > Multivariate analysis > Correspondence analysis >
Postestimation after MCA or JCA > Dimension projection plot
Options for mcaprojection
+-------+
----+ Plots +------------------------------------------------------------
plot_options affect the rendition of markers, including their shape,
size, color, and outline (see [G] marker_options) and specify if and
how the markers are to be labeled (see [G] marker_label_options).
These options may be specified for each variable. If the overlay
option is not specified then for each variable you may also specify
twoway_options excluding by() and name().
+---------+
----+ Options +----------------------------------------------------------
dimensions(numlist) identifies the dimensions to be displayed. By
default all dimensions are displayed.
normalize(norm) specifies the normalization of the coordinates.
normalize(standard) returns coordinates in standard normalization.
normalize(principal) returns principal coordinates. The default is
the normalization method specified with mca during estimation, or
normalize(standard) if no method was specified.
alternate causes adjacent labels to alternate sides.
maxlength(#) specifies the maximum number of characters for row and
column labels; the default is maxlength(12).
combine_options affect the rendition of the combined plot; see [G] graph
combine. These options may not be used if only one variable is
specified.
+-----------------------------------------+
----+ Y axis, X axis, Titles, Legend, Overall +--------------------------
twoway_options are any of the options documented in [G] twoway_options,
excluding by().
Examples
Setup
. webuse issp93
. mca A B C D, dimensions(2) suppl(age edu) method(joint)
Predict column coordinates and row coordinates
. predict a1 a2, score(A)
. predict r1 r2, rowscores norm(principal)
View the coordinates and the subinertia
. estat coord, stats
. estat subinertia
Biplots
. mcaplot
. mcaplot A B C, ynegate
. mcaplot (A, mcolor(red) mlabcolor(red)) (B, mcolor(blue)), overlay
Dimension projection plots
. mcaprojection
. mcaprojection A B C, alternate
Saved results
estat summarize saves the following in r():
Matrices
r(stats) k x 4 matrix of means, standard deviations,
minimums, and maximums
estat coordinates saves the following in r()
Macros
r(norm) the normalization method of the coordinates
Matrices
r(Coord) column coordinates
r(Stats) column statistics: mass, distance, and inertia
(option stats only)
estat subinertia saves the following in r():
Matrices
r(inertia_sub) variable-by-variable inertias
Also see
Manual: [MV] mca postestimation
Help: [MV] mca; [MV] ca, [MV] ca postestimation
