Привет всем.Я застрял с простым кодом.На самом деле у меня есть такой вывод:
roc_value
, , MAXENT.Phillips, RUN1, PA1
Testing.data Cutoff Sensitivity Specificity
KAPPA 0.558 484.0 63.713 90.667
TSS 0.555 354.0 73.840 81.667
ROC 0.863 369.5 72.996 83.000
ACCURACY 0.788 484.0 63.713 90.667
BIAS 0.996 344.0 73.840 79.667
, , GLM, RUN1, PA1
Testing.data Cutoff Sensitivity Specificity
KAPPA 0.565 422 83.122 74.333
TSS 0.575 422 83.122 74.333
ROC 0.843 420 83.122 74.333
ACCURACY 0.782 422 83.122 74.333
BIAS 0.979 518 72.574 79.667
, , GAM, RUN1, PA1
Testing.data Cutoff Sensitivity Specificity
KAPPA 0.664 554.0 81.435 85.000
TSS 0.664 554.0 81.435 85.000
ROC 0.885 557.5 81.435 85.333
ACCURACY 0.834 595.0 76.793 88.667
BIAS 0.996 554.0 81.435 85.000
, , RF, RUN1, PA1
Testing.data Cutoff Sensitivity Specificity
KAPPA 0.768 404 83.544 92.000
TSS 0.762 404 83.544 92.000
ROC 0.950 406 83.544 92.667
ACCURACY 0.886 485 81.435 94.333
BIAS 0.987 343 84.810 89.000
, , MAXENT.Tsuruoka, RUN1, PA1
Testing.data Cutoff Sensitivity Specificity
KAPPA 0.509 368 86.498 66.000
TSS 0.527 321 90.717 61.667
ROC 0.838 372 86.498 66.333
ACCURACY 0.756 479 74.262 76.333
BIAS 0.970 488 70.464 78.667
str (roc_value):
num [1:5, 1:4, 1:5, 1, 1] 0.558 0.555 0.863 0.788 0.996 ...
- attr(*, "dimnames")=List of 5
..$ : chr [1:5] "KAPPA" "TSS" "ROC" "ACCURACY" ...
..$ : chr [1:4] "Testing.data" "Cutoff" "Sensitivity" "Specificity"
..$ : chr [1:5] "MAXENT.Phillips" "GLM" "GAM" "RF" ...
..$ : chr "RUN1"
..$ : Named chr "PA1"
.. ..- attr(*, "names")= chr ""
Я хочу выбрать только значения ROC из столбца Testing.data в одном объекте (дляпродолжить мой сценарий), например:
whatidlike <- c(0.863,0.843,0.885,0.950,0.838) # only a single object with ROC values.
Пробовал dplyr
, subset
и т. д. и не может двигаться дальше.Может ли кто-нибудь помочь мне?