.BG
.VE $Header: /usr/s/current/s/.help/RCS/regsum,v 1.3 84/11/05 11:17:05 rab Exp $
.FN regsum
.TL
regsum: Regression Summaries
.CS
regsum(z)
.PP
.AG z
regression structure, with components `coef', `resid', `r',
`corth', `int' and (possibly) `sqrtw'.  See documentation
for `regress'.
.RT
structure with components summarizing the regression
statistics.
.RC names
vector of names for the `x' variables in the regression:
either the names provided as arguments to `reg' or "x1", "x2", etc.
If the regression has an intercept, "Intercept" is
the first name.
.RC ynames
vector of names for the `y' variables in the regression:
either the names provided as arguments to `reg' or "y1", "y2", etc.
.RC coef
coefficient matrix.
.RC stderr
matrix of standard errors associated with the
coefficients of each regression.
.RC t
matrix of t-statistics associated with the coefficients
of each regression.
.RC cor
correlation matrix of the coefficients.
.RC cov
covariance matrix of the coefficients.
If there is more than one y variable, component `stdcov' is returned instead of `cov'.
.RC stdcov
standardized covariance matrix such that
the actual covariance matrix corresponding to each y variable is the
standardized covariance matrix times the square of the
corresponding `rms' value.
If there is only one y variable, component `cov' is returned instead of `stdcov'.
.RC rsq
vector of multiple R-square statistics (the fraction of the variance of
`y' explained by the model).
.RC rms
vector of standard errors of the residuals.
.RC fval
vector of the overall F-statistics for the regressions.
.RC df
the degrees of freedom for numerator and denominator of `fval'.
.EX
z <- reg(x,y) #do the regression
zsumy <- regsum(z) #get the summary
.KW regression
.KW array*
.WR
