LowCon: A Design-based Subsampling Approach in a Misspecified Linear Model

A subsampling method based space-filling design for misspecified linear model proposed by Meng et.al. (2021).

Usage

LowCon(X, n, theta = 1, space_method = "randomLHS", seed = NULL)

Arguments

X: A data.frame or matrix of explanatory variables.
n: Subsample size.
theta: Percentage of data shrinkage. Default to 1.
space_method: The generation method of initial space-filling design.
seed: Random seed for the sampling.

Value

Subsample index.

References

Cheng Meng, Rui Xie, Abhyuday Mandal, Xinlian Zhang, Wenxuan Zhong & Ping Ma (2021) LowCon: A Design-based Subsampling Approach in a Misspecified Linear Model, Journal of Computational and Graphical Statistics, 30:3, 694-708, https://www.tandfonline.com/doi/full/10.1080/10618600.2020.1844215.

Examples

data <- data_numeric_regression
X <- data[-which(names(data) == "y")]
LowCon(X, n = 100, space_method = "randomLHS", theta = 1, seed = NULL)
#>   [1] 6210 7368 7559 5199 5116  445 8331 1262  704 7416 9560 5289 2345 5992 7002
#>  [16] 5072 8578 9090 4328 2938 1161 4703 1026 3912 7409 2574 3183 3551 1591 3931
#>  [31] 9536 6006 1186 2198  741 7594 7359 2211 2397 5523 4696 6114 3162 9337 1839
#>  [46] 5589 6148 6677 9027 3129 4101 4983 1456 1038 2456 3545 9164  830 3246 5825
#>  [61] 8444 6591 9594 1804 2831 9529 9661  138 8257 7237 5488 5560 6992 2875 1667
#>  [76] 5221  334   30 2206 1830  658 8944 5831 5564 9904 2334 9096  690 4612 2434
#>  [91] 9902 3175 1794 7053 9472 9017 3226 3444  260 7073