Summary method for cv.ncvreg
objects
A "cv.ncvreg"
or "cv.ncvsurv"
object.
Further arguments passed to or from other methods.
A "summary.cv.ncvreg"
object.
Number of digits past the decimal point to print out. Can be a vector specifying different display digits for each of the five non-integer printed values.
summary.cv.ncvreg
produces an object with S3 class
"summary.cv.ncvreg"
. The class has its own print method and contains
the following list elements:
The penalty used by
ncvreg
.
Either "linear"
or "logistic"
,
depending on the family
option in ncvreg
.
Number of observations
Number of regression coefficients (not including the intercept).
The index of lambda
with the smallest
cross-validation error.
The sequence of lambda
values
used by cv.ncvreg
.
Cross-validation error (deviance).
Proportion of variance explained by the model, as estimated by cross-validation. For models outside of linear regression, the Cox-Snell approach to defining R-squared is used.
Signal to noise ratio, as estimated by cross-validation.
For linear regression models, the scale parameter estimate.
For logistic regression models, the prediction error (misclassification error).
Breheny P and Huang J. (2011) Coordinate descentalgorithms for nonconvex penalized regression, with applications to biological feature selection. Annals of Applied Statistics, 5: 232-253. c("\Sexpr[results=rd]tools:::Rd_expr_doi(\"#1\")", "10.1214/10-AOAS388")doi:10.1214/10-AOAS388
# Linear regression --------------------------------------------------
data(Prostate)
cvfit <- cv.ncvreg(Prostate$X, Prostate$y)
summary(cvfit)
#> MCP-penalized linear regression with n=97, p=8
#> At minimum cross-validation error (lambda=0.0224):
#> -------------------------------------------------
#> Nonzero coefficients: 7
#> Cross-validation error (deviance): 0.54
#> R-squared: 0.59
#> Signal-to-noise ratio: 1.46
#> Scale estimate (sigma): 0.732
#> MCP-penalized linear regression with n=97, p=8
#> At lambda=0.0224:
#> -------------------------------------------------
#> Nonzero coefficients : 7
#> Expected nonzero coefficients: 2.50
#> Average mfdr (7 features) : 0.357
#>
#> Estimate z mfdr Selected
#> lcavol 0.569546 8.986 < 1e-04 *
#> svi 0.752398 4.170 0.00080455 *
#> lweight 0.614420 3.524 0.00822827 *
#> pgg45 0.005324 2.010 0.48176729 *
#> lbph 0.097353 1.891 0.61027217 *
#> age -0.020913 -2.084 0.62413088 *
#> lcp -0.104959 -1.965 0.77570294 *
# Logistic regression ------------------------------------------------
data(Heart)
cvfit <- cv.ncvreg(Heart$X, Heart$y, family="binomial")
summary(cvfit)
#> MCP-penalized logistic regression with n=462, p=9
#> At minimum cross-validation error (lambda=0.0270):
#> -------------------------------------------------
#> Nonzero coefficients: 5
#> Cross-validation error (deviance): 1.07
#> R-squared: 0.20
#> Signal-to-noise ratio: 0.25
#> Prediction error: 0.279
#> MCP-penalized logistic regression with n=462, p=9
#> At lambda=0.0270:
#> -------------------------------------------------
#> Nonzero coefficients : 5
#> Expected nonzero coefficients: 0.06
#> Average mfdr (5 features) : 0.011
#>
#> Estimate z mfdr Selected
#> age 0.05109 5.946 < 1e-04 *
#> famhist 0.90619 4.143 0.00059229 *
#> tobacco 0.07012 3.328 0.01113128 *
#> typea 0.03045 3.169 0.01861106 *
#> ldl 0.13459 3.062 0.02605639 *
# Cox regression -----------------------------------------------------
data(Lung)
cvfit <- cv.ncvsurv(Lung$X, Lung$y)
summary(cvfit)
#> MCP-penalized Cox regression with n=137, p=8
#> At minimum cross-validation error (lambda=0.1573):
#> -------------------------------------------------
#> Nonzero coefficients: 3
#> Cross-validation error (deviance): 7.53
#> R-squared: 0.29
#> Signal-to-noise ratio: 0.42
#> MCP-penalized Cox regression with n=137, p=8
#> At lambda=0.1573:
#> -------------------------------------------------
#> Nonzero coefficients : 3
#> Expected nonzero coefficients: 0.26
#> Average mfdr (3 features) : 0.086
#>
#> Estimate z mfdr Selected
#> karno -0.03322 -6.564 < 1e-04 *
#> squamous -0.31025 -2.872 0.10842 *
#> adeno 0.18197 2.689 0.14969 *