weighted.lasso.computations: Wrapper function to store results for weighted lasso variable...
In rakheon/d2wlasso: Structured variable selection with weighted lasso

Description Usage Arguments Value References

Performs variable selection with covariates multiplied by weights that direct which variables are likely to be associated with the response.

weighted.lasso.computations(
  weights,
  weight_fn = function(x) {     x },
  response,
  XX,
  z,
  z.names,
  show.plots,
  penalty.choice,
  est.MSE,
  cv.folds,
  delta
)

`weights`	(m x 1) matrix that we use to multiply the m-covariates by.
`weight_fn`	A user-defined function to be applied to the weights for the weighted lasso. Default is an identify function.
`response`	a list containing: yy and delta. yy is an (n by 1) matrix corresponding to the response variable. If `regression.type` is "cox", `y` contains the observed event times. delta is an (n by 1) matrix that denotes censoring when `regression.type` is "cox" (1 denotes survival event is observed, 0 denotes the survival event is censored). Can be NULL.
`XX`	(n by K) matrix of main covariates where n is the sample size and K=m if z is NULL, and K= m+1 otherwise. Here, m refers to the number of x-covariates.
`z`	(n by 1) matrix of additional fixed covariate affecting response variable. This covariate will always be selected. Can be NULL.
`z.names`	character denoting the column name of the z-covariate if z is not NULL. Can be NULL.
`show.plots`	logical indicator. If TRUE and `penalty.choice` is "penalized.loss", a plot of the penalized loss criterion versus steps in the LARS algorithm of Efron et al (2004).
`penalty.choice`	character that indicates the variable selection criterion. Options are "cv.mse" for the K-fold cross-validated mean squared prediction error, "penalized.loss" for the penalized loss criterion which requires specification of the penalization parameter `penalized.loss.delta`, "cv.penalized.loss" for the K-fold cross-validated criterion to determine delta in the penalized loss criterion, and "deviance.criterion" for optimizing the Cox proportional hazards deviance (only available when `regression.type` is "cox".) Defalt is "penalized.loss".
`est.MSE`	character that indicates how the mean squared error is estimated in the penalized loss criterion when `penalty.choice` is "penalized.loss" or "cv.penalized.loss". Options are "est.var" which means the MSE is sd(y) * sqrt(n/(n-1)) where n is the sample size, and "step" which means we use the MSE from forward stepwise regression with AIC as the selection criterion. Default is "est.var".
`cv.folds`	scalar denoting the number of folds for cross-validation when `penalty.choice` is "cv.mse" or "cv.penalized.loss". Default is 10.
`delta`	scalar to indicate the choice of the penalization parameter delta in the penalized loss criterion when `penalty.choice` is "penalized.loss".

interest:(m by 1) matrix where the kth entry is 1 if x_k is selected to be in the model, and 0 if not.
sign.interest:(m by 1) matrix where the kth entry is 1 if the coefficient in front of x_k is positive, and -1 if not.
delta.out:the penalization parameter delta used in the penalized loss criterion. When penalty.choice is "penalized.loss", delta.out is the same as delta. When penalty.choice is "cv.penalized.loss", delta.out is the resulting delta-value obtained from the k-fold cross validation.

Efron, B., Hastie, T., Johnstone, I. AND Tibshirani, R. (2004). Least angle regression. Annals of Statistics 32, 407–499.

Garcia, T.P. and M¨uller, S. (2016). Cox regression with exclusion frequency-based weights to identify neuroimaging markers relevant to Huntington’s disease onset. Annals of Applied Statistics, 10, 2130-2156.

Garcia, T.P. and M¨uller, S. (2014). Influence of measures of significance-based weights in the weighted Lasso. Journal of the Indian Society of Agricultural Statistics (Invited paper), 68, 131-144.

Garcia, T.P., Mueller, S., Carroll, R.J., Dunn, T.N., Thomas, A.P., Adams, S.H., Pillai, S.D., and Walzem, R.L. (2013). Structured variable selection with q-values. Biostatistics, DOI:10.1093/biostatistics/kxt012.

Storey, J. D. and Tibshirani, R. (2003). Statistical significance for genomewide studies. Proceedings of the National Academy of Sciences 100, 9440-9445.

rakheon/d2wlasso documentation built on Feb. 26, 2020, 10:39 p.m.