Compute Diagnostics for `lsfit' Regression Results
Usage
ls.diag(ls.out)
Arguments
ls.out
|
Typically the result of lsfit()
|
Description
Computes basic statistics, including standard errors, t- and p-values
for the regression coefficients.Value
A list
with the following numeric components.
std.dev
|
The standard deviation of the errors, an estimate of
sigma.
|
hat
|
diagonal entries h_{ii} of the hat matrix H
|
std.res
|
standardized residuals
|
stud.res
|
studentized residuals
|
cooks
|
Cook's distances
|
dfits
|
DFITS statistics
|
correlation
|
correlation matrix
|
std.err
|
standard errors of the regression coefficients
|
cov.scaled
|
Scaled covariance matrix of the coefficients
|
cov.unscaled
|
Unscaled covariance matrix of the coefficients
|
References
Belsley, D. A., E. Kuh and R. E. Welsch (1980).
Regression Diagnostics.
New York: Wiley.See Also
hat
for the hat matrix diagonals,
ls.print
,
lm.influence
, summary.lm
,
anova
.Examples
##-- Using the same data as the lm(.) example:
lsD9 <- lsfit(x = as.numeric(gl(2, 10, 20)), y = weight)
dlsD9 <- ls.diag(lsD9)
str(dlsD9, give.attr=FALSE)
abs(1 - sum(dlsD9$hat) / 2) < 10*.Machine$double.eps # sum(h.ii) = p
plot(dlsD9$hat, dlsD9$stud.res, xlim=c(0,0.11))
abline(h = 0, lty = 2, col = "lightgray")
##-- Using the same data as the lm(.) example:
lsD9 <- lsfit(x = as.numeric(gl(2, 10, 20)), y = weight)
dlsD9 <- ls.diag(lsD9)
str(dlsD9, give.attr=FALSE)
abs(1 - sum(dlsD9$hat) / 2) < 10*.Machine$double.eps # sum(h.ii) = p
plot(dlsD9$hat, dlsD9$stud.res, xlim=c(0,0.11))
abline(h = 0, lty = 2, col = "lightgray")