Skip to contents

A model-free subsampling method based on uniform design proposed by Zhang et.al. (2023).

Usage

DDS(X, n, ratio = 0.85)

Arguments

X

A data.frame or matrix of explanatory variables.

n

Subsample size.

ratio

Dimensionality reduction ratio of PCA, default to 0.85.

Value

Subsample index.

Details

The uniform design is generated according to the method described in part 5 of Zhang's article. The uniformity measure uses mixed divergence.

References

Mei Zhang, Yongdao Zhou, Zheng Zhou & Aijun Zhang (2023) Model-Free Subsampling Method Based on Uniform Designs, IEEE Transactions on Knowledge and Data Engineering, 36:3, 1210-1220, https://ieeexplore.ieee.org/abstract/document/10192374.

Examples

data <- data_numeric_regression
X <- data[-which(names(data) == "y")]
DDS(X, n = 100, ratio = 0.85)
#>      [,1] [,2] [,3] [,4] [,5] [,6] [,7] [,8] [,9] [,10] [,11] [,12] [,13] [,14]
#> [1,] 2238 2843 7010 8283 9340 1738  787 9555 1342  4120  9243   718   688  4792
#>      [,15] [,16] [,17] [,18] [,19] [,20] [,21] [,22] [,23] [,24] [,25] [,26]
#> [1,]  3111  5176  6981  3831  4353  9894  3000  7966   311  8780  9324  1742
#>      [,27] [,28] [,29] [,30] [,31] [,32] [,33] [,34] [,35] [,36] [,37] [,38]
#> [1,]  1390  8222  1827  7583  2335  9181  3109  2041  2395  6539  7812  5616
#>      [,39] [,40] [,41] [,42] [,43] [,44] [,45] [,46] [,47] [,48] [,49] [,50]
#> [1,]  1753  5028  3147   460  7741  5427   670  3965   824  4025   406  9813
#>      [,51] [,52] [,53] [,54] [,55] [,56] [,57] [,58] [,59] [,60] [,61] [,62]
#> [1,]  8686  6071    87  8553    85  2886   492  6377  3126  3064   423  8448
#>      [,63] [,64] [,65] [,66] [,67] [,68] [,69] [,70] [,71] [,72] [,73] [,74]
#> [1,]   611  2565  6673  1971  9727  5396  3745  7576  4819  8385  4268  5835
#>      [,75] [,76] [,77] [,78] [,79] [,80] [,81] [,82] [,83] [,84] [,85] [,86]
#> [1,]  6308  5660  7875  4666  4072  3578  2771  1801  4466  5332  4813  9383
#>      [,87] [,88] [,89] [,90] [,91] [,92] [,93] [,94] [,95] [,96] [,97] [,98]
#> [1,]  1443  3293  4392  4804  2998     8  2750  6135  5014   657  5944  9737
#>      [,99] [,100]
#> [1,]  2651   2354