Feature selection using XGBoost

Usage

ML_xgboost(object)

Arguments

object: A dataframe-like data object containing log-metabolite intensity values, with columns corresponding to metabolites and must containing the group column, and the rows corresponding to the samples

Value

test

Examples

library(dplyr)

meta_dat1 <- t(meta_dat) %>%
  as.data.frame() %>%
  dplyr::mutate(group=group)
result_ML_xgboost <- ML_xgboost(meta_dat1)
#> [1]	train-rmse:0.361110	test-rmse:0.472223 
#> [2]	train-rmse:0.260802	test-rmse:0.476949 
#> [3]	train-rmse:0.188357	test-rmse:0.493187 
#> [4]	train-rmse:0.136036	test-rmse:0.511018 
#> [5]	train-rmse:0.098249	test-rmse:0.526761 
#> [6]	train-rmse:0.070959	test-rmse:0.539494 
#> [7]	train-rmse:0.051248	test-rmse:0.549350 
#> [8]	train-rmse:0.037013	test-rmse:0.556795 
#> [9]	train-rmse:0.026732	test-rmse:0.562335 
#> [10]	train-rmse:0.019307	test-rmse:0.566419 
#> [11]	train-rmse:0.013944	test-rmse:0.569410 
#> [12]	train-rmse:0.010071	test-rmse:0.571592 
#> [13]	train-rmse:0.007274	test-rmse:0.573179 
#> [14]	train-rmse:0.005253	test-rmse:0.574332 
#> [15]	train-rmse:0.003794	test-rmse:0.575167 
#> [16]	train-rmse:0.002740	test-rmse:0.575772 
#> [17]	train-rmse:0.001979	test-rmse:0.576209 
#> [18]	train-rmse:0.001430	test-rmse:0.576526 
#> [19]	train-rmse:0.001032	test-rmse:0.576754 
#> [20]	train-rmse:0.000746	test-rmse:0.576920 
#> [21]	train-rmse:0.000539	test-rmse:0.577039 
#> [22]	train-rmse:0.000389	test-rmse:0.577126 
#> [23]	train-rmse:0.000281	test-rmse:0.577188 
#> [24]	train-rmse:0.000203	test-rmse:0.577233 
#> [25]	train-rmse:0.000203	test-rmse:0.577233 
#> [26]	train-rmse:0.000203	test-rmse:0.577233 
#> [27]	train-rmse:0.000203	test-rmse:0.577233 
#> [28]	train-rmse:0.000203	test-rmse:0.577233 
#> [29]	train-rmse:0.000203	test-rmse:0.577233 
#> [30]	train-rmse:0.000203	test-rmse:0.577233 
#> [31]	train-rmse:0.000203	test-rmse:0.577233 
#> [32]	train-rmse:0.000203	test-rmse:0.577233 
#> [33]	train-rmse:0.000203	test-rmse:0.577233 
#> [34]	train-rmse:0.000203	test-rmse:0.577233 
#> [35]	train-rmse:0.000203	test-rmse:0.577233 
#> [36]	train-rmse:0.000203	test-rmse:0.577233 
#> [37]	train-rmse:0.000203	test-rmse:0.577233 
#> [38]	train-rmse:0.000203	test-rmse:0.577233 
#> [39]	train-rmse:0.000203	test-rmse:0.577233 
#> [40]	train-rmse:0.000203	test-rmse:0.577233 
#> [41]	train-rmse:0.000203	test-rmse:0.577233 
#> [42]	train-rmse:0.000203	test-rmse:0.577233 
#> [43]	train-rmse:0.000203	test-rmse:0.577233 
#> [44]	train-rmse:0.000203	test-rmse:0.577233 
#> [45]	train-rmse:0.000203	test-rmse:0.577233 
#> [46]	train-rmse:0.000203	test-rmse:0.577233 
#> [47]	train-rmse:0.000203	test-rmse:0.577233 
#> [48]	train-rmse:0.000203	test-rmse:0.577233 
#> [49]	train-rmse:0.000203	test-rmse:0.577233 
#> [50]	train-rmse:0.000203	test-rmse:0.577233 
#> [51]	train-rmse:0.000203	test-rmse:0.577233 
#> [52]	train-rmse:0.000203	test-rmse:0.577233 
#> [53]	train-rmse:0.000203	test-rmse:0.577233 
#> [54]	train-rmse:0.000203	test-rmse:0.577233 
#> [55]	train-rmse:0.000203	test-rmse:0.577233 
#> [56]	train-rmse:0.000203	test-rmse:0.577233 
#> [57]	train-rmse:0.000203	test-rmse:0.577233 
#> [58]	train-rmse:0.000203	test-rmse:0.577233 
#> [59]	train-rmse:0.000203	test-rmse:0.577233 
#> [60]	train-rmse:0.000203	test-rmse:0.577233 
#> [61]	train-rmse:0.000203	test-rmse:0.577233 
#> [62]	train-rmse:0.000203	test-rmse:0.577233 
#> [63]	train-rmse:0.000203	test-rmse:0.577233 
#> [64]	train-rmse:0.000203	test-rmse:0.577233 
#> [65]	train-rmse:0.000203	test-rmse:0.577233 
#> [66]	train-rmse:0.000203	test-rmse:0.577233 
#> [67]	train-rmse:0.000203	test-rmse:0.577233 
#> [68]	train-rmse:0.000203	test-rmse:0.577233 
#> [69]	train-rmse:0.000203	test-rmse:0.577233 
#> [70]	train-rmse:0.000203	test-rmse:0.577233 


result_ML_xgboost$p

result_ML_xgboost$feature_result
#>    Feature  Gain Cover Frequency Importance
#>     <fctr> <num> <num>     <num>      <num>
#> 1:  C00267     1     1         1          1