Title
leanout -- Lean output formatting for regressions results
Syntax
Standard estimation command syntax
leanout [, options ] : estimation_command
Survey estimation command syntax
leanout [, options ] : svy [vcetype] [, svy_options ] : estimation_command
Where estimation_command is almost any estimation command and includes any options for the estimation command.
options Description ------------------------------------------------------------------------- Reporting format(fmt) display format format for results varwidth(#) allow # characters for displaying variable names
display_options control spacing and display of omitted variables and base and empty cells ------------------------------------------------------------------------- Weights are allowed if command_name allows them; see weight. All postestimation commands behave as they would after the estimation command without the leanout prefix. leanout: [, options ] may be typed to replay the estimation results leanout: [, options ] may be typed after normal estimation with most estimation commands to display the lean output.
Description
leanout fits a model specified by estimation_command. It suppresses the command's normal output and instead displays a "lean" estimation table comprised of only the variable names, the coefficients, the standard errors, and the confidence intervals.
Options
format(fmt) specifies the display format used to display the coefficient and its standard error and confidence interval. Default format is %f5.1f.
varwidth(#) specifies the maximum number of characters for displaying variable names. Names longer than # are abbreviated. Default is 16 characters.
display_options may be noomitted, vsquish, noemptycells, baselevels, allbaselevels; see [R] estimation options.
Remarks
leanout works with almost all estimation commands. It does not produce useful output after exlogistic, expoisson, asmprobit, asclogit, asroprobit.
leanout does not support alternate parameterization of estimated coefficients, such as odds ratios for logistic models.
Most maximim-likelihood estimators, and other estimators requiring optimization, estimate ancillary parameters in a transformed metric to improve convergence. leanout displays ancillary parameters only in the metric in which they were estimated. It does not display them in their natural, untransformed metric.
Examples
Linear regression . sysuse auto . leanout: regress length turn headroom
Replay results . leanout:
Linear regression with factor variables . leanout: regress length turn headroom i.for i.rep78
Showing base categories . leanout, base: regress length turn headroom i.for i.rep78
With a custom format . leanout, format(%8.4f): regress length turn headroom i.for i.rep78
As a postestimation command . regress length turn headroom i.for i.rep78 . leanout:
Linear regression with survey data . webuse nhanes2f: . svyset psuid [pweight=finalwgt], strata(stratid): . leanout: svy: regress zinc age female black orace rural:
A maximum-likelihood estimator . webuse lbw . leanout, format(%5.2f): probit low smoke ptl ht ui i.race
A multiple equation estimator . sysuse auto . leanout: sureg (mpg trunk length) (turn trunk i.foreign)
Saved results
leanout leaves behind the estimation command's standard saved results in e():
Also see
Help: [R] estimates table