help estimates table dialog: estimates table
-------------------------------------------------------------------------------
Title
[R] estimates table -- Compare estimation results
Syntax
estimates table [namelist] [, options]
where namelist is a name, a list of names, _all, or *.
A name may be ., meaning the current (active) estimates.
_all and * mean the same thing.
options description
-------------------------------------------------------------------------
Main
stats(scalarlist) report scalarlist in table
star[(#1 #2 #3)] use stars to denote significance levels
Options
keep(coeflist) report coefficients in order specified
drop(coeflist) omit specified coefficients from table
equations(matchlist) match equations of models as specified
Numerical formats
b[(%fmt)] how to format coefficients, which are always
reported
se[(%fmt)] report standard errors and use optional format
t[(%fmt)] report t or z and use optional format
p[(%fmt)] report p-values and use optional format
stfmt(%fmt) how to format scalar statistics
General format
varwidth(#) use # characters to display variable names and
statistics
modelwidth(#) use # characters to display model names
eform display coefficients in exponentiated form
label display variable labels rather than variable
names
newpanel display statistics in separate table from
coefficients
style(oneline) put vertical line after variable names; the
default
style(columns) put vertical line separating every column
style(noline) suppress all vertical lines
coded display compact table
Reporting
display_options control spacing and display of omitted variables
and base and empty cells
+ title(string) title for table
-------------------------------------------------------------------------
+ title() does not appear in the dialog box.
where
o A scalarlist is a list of any or all of the names of scalars stored
in e(), plus aic, bic, and rank.
o #1 #2 #3 are three numbers such as .05 .01 .001.
o A coeflist is a list of coefficient names, each name of which may
be simple (e.g., price), an equation name followed by a colon
(e.g., mean:), or a full name (e.g., mean:price). Names are
separated by blanks.
o A matchlist specifies how equations from different estimation
results are to be matched. If you need to specify a matchlist,
the solution is usually 1, as in equations(1). The full syntax
is
matchlist := term [, term ...]
term := [eqname =] #:#...:#
[eqname =] #
See equations() under Options below.
o %fmt is any valid Stata numerical display format.
Menu
Statistics > Postestimation > Manage estimation results > Table of fit
statistics
Description
estimates table displays a table of coefficients and statistics for one
or more sets of estimation results.
Options
+------+
----+ Main +-------------------------------------------------------------
stats(scalarlist) specifies one or more scalar statistics to be displayed
in the table. scalarlist may contain
aic Akaike's information criterion
bic Schwarz's Bayesian information criterion
rank rank of e(V) (# of free parameters in model)
along with the names of any scalars stored in e(). The specified
statistics do not have to be available for all estimation results
being displayed.
For example, stats(N ll chi2 aic) specifies that e(N), e(ll),
e(chi2), and AIC be included. In Stata, e(N) records the number of
observations; e(ll), the log likelihood; and e(chi2), the chi-squared
test that all coefficients in the first equation of the model are
equal to zero.
star and star(#1 #2 #3) specify that stars (asterisks) are to be used to
mark significance. The second syntax specifies the significance
levels for one, two, and three stars. If you specify simply star,
that is equivalent to specifying star(.05 .01 .001), which means one
star (*) if p < 0.05, two stars (**) if p < 0.01, and three stars
(***) if p < 0.001.
The star and star() options may not be combined with the se, t, or p
options.
+---------+
----+ Options +----------------------------------------------------------
keep(coeflist) and drop(coeflist) are alternatives; they specify
coefficients to be included or omitted from the table. The default
is to display all coefficients.
If keep() is specified, it specifies not only the coefficients to be
included but also the order in which they appear.
A coeflist is a list of coefficient names, each name of which may be
simple (e.g., price), an equation name followed by a colon (e.g.,
mean:), or a full name (e.g., mean:price). Names are separated from
each other by blanks.
When full names are not specified, all coefficients that match the
partial specification are included. For instance, drop(_cons) would
omit _cons for all equations.
equations(matchlist) specifies how the equations of the models in
namelist are to be matched. The default is to match equations by
name. Matching by name usually works well when all results were fit
by the same estimation command. When you are comparing results from
different estimation commands, however, specifying equations() may be
necessary.
The most common usage is equations(1), which indicates that all first
equations are to be matched into one equation named #1.
matchlist has the syntax
term [, term ... ]
where term is
[eqname =] #:#...:# (syntax 1)
[eqname =] # (syntax 2)
In syntax 1, each # is a number or a period (.). If a number, it
specifies the position of the equation in the corresponding model;
1:3:1 would indicate that equation 1 in the first model matches
equation 3 in the second, which matches equation 1 in the third. A
period indicates that there is no corresponding equation in the
model; 1:.:1 indicates that equation 1 in the first matches equation
1 in the third.
In syntax 2, you specify just one number, say, 1 or 2, and that is
shorthand for 1:1...:1 or 2:2...:2, meaning that equation 1 matches
across all models specified or that equation 2 matches across all
models specified.
Now that you can specify a term, you can put that together into a
matchlist by separating one term from the other by commas. In what
follows, we will assume that three names were specified,
. estimates table alpha beta gamma, ...
equations(1) is equivalent to equations(1:1:1); we would be saying
that the first equations match across the board.
equations(1:.:1) would specify that equation 1 matches in models
alpha and gamma but that there is nothing corresponding in model
beta.
equations(1,2) is equivalent to equations(1:1:1, 2:2:2). We would be
saying that the first equations match across the board and so do the
second equations.
equations(1, 2:.:2) would specify that the first equations match
across the board, that the second equations match for models alpha
and gamma, and that there is nothing equivalent to equation 2 in
model beta.
If equations() is specified, equations not matched by position are
matched by name.
+-------------------+
----+ Numerical formats +------------------------------------------------
b(%fmt) specifies how the coefficients are to be displayed. You might
specify b(%9.2f) to make decimal points line up. There is also a b
option, which specifies that coefficients are to be displayed, but
that is just included for consistency with the se, t, and p options.
Coefficients are always displayed.
se, t, and p specify that standard errors, t or z statistics, and
significance levels are to be displayed. The default is not to
display them. se(%fmt), t(%fmt), and p(%fmt) specify that each is to
be displayed and specifies the display format to be used to format
them.
stfmt(%fmt) specifies the format for displaying the scalar statistics
included by the stats() options.
+----------------+
----+ General format +---------------------------------------------------
varwidth(#) specifies the number of character positions used to display
the names of the variables and statistics. The default is 12.
modelwidth(#) specifies the number of character positions used to display
the names of the models. The default is 12.
eform displays coefficients in exponentiated form. For each coefficient,
exp(b) rather than b is displayed, and standard errors are
transformed appropriately. Display of the intercept, if any, is
suppressed.
label specifies that variable labels be displayed instead of variable
names.
newpanel specifies that the statistics be displayed in a table separated
by a blank line from the table with coefficients rather than in the
style of another equation in the table of coefficients.
style(stylespec) specifies the style of the coefficient table.
style(oneline) specifies that a vertical line be displayed after the
variables but not between the models. This is the default.
style(columns) specifies that vertical lines be displayed after each
column.
style(noline) specifies that no vertical lines be displayed.
coded specifies that a compact table be displayed. This format is
especially useful for comparing variables that are included in a
large collection of models.
+-----------+
----+ Reporting +--------------------------------------------------------
display_options: noomitted, vsquish, noemptycells, baselevels,
allbaselevels; see [R] estimation options.
The following option is available with estimates table but is not shown
in the dialog box:
title(string) specifies the title to appear above the table.
Examples
Setup
. sysuse auto
. regress mpg gear turn
. estimates store small
. regress mpg gear turn length
. estimates store large
Create table of results, displaying sample size and adjusted R-squared in
addition to coefficients and standard errors
. estimates table small large, b(%7.4f) se(%7.4f) stats(N r2_a)
Same as above, but use a format for the scalar statitics that will show
the samples size without a decimal point
. estimates table large, b(%7.4f) se(%7.4f) stfmt(%7.4g) stats(N
r2_a)
Saved results
estimates table saves the following in r():
Macros
r(names) names of results used
Matrices
r(coef) matrix M: n x 2*m
M[i, 2j-1] = ith parameter estimate for model j;
M[i, 2j ] = variance of M[i, 2j-1];
i = 1, ..., n; j = 1, ..., m
r(stats) matrix S: k x m (if option stats() specified)
S[i, j] = ith statistic for model j;
i = 1, ..., k; j = 1, ..., m
Also see
Manual: [R] estimates table
Help: [R] estimates
