restricted ROC

Calculate the restricted ROC curves.

Usage

simple_rROC(
  response,
  predictor,
  direction = "<",
  positive_label = NULL,
  get_all_aucs_fun = get_all_aucs_norecalculation,
  return_proc = FALSE,
  do_parallel = FALSE,
  check_positive_negative_count = FALSE
)

Arguments

response

A vector containing the true class labels. Care that it is VERY important which class is the positive class because the predictions are ordered according to restriction

predictor

A vector containing the predictions.

direction

See pROC::roc(), but only "<" is implemented right now. Maybe changing the positive_label already solves your problem.

positive_label

Label for the positive class. All other values of response are regarded as negative cases.

get_all_aucs_fun

How to calculate the AUCs. You would usually now want to set that. Implemented are get_all_aucs(): Calculates the AUCs by actively splitting the data into markerHIGH and markerLOW parts. Then calculates a usual AUC on the parts.

    `get_all_aucs_norecalculation()`:
        Calculates the AUCs based on the scaling factor described in the
        publication. Much faster after the ROC curve does not have to
        be recalculated over and over again.
        Todo: Could potentially be improved by not recalculating the partial
        AUCs with pROC over and over but by just adding parts.

return_proc

1. Should pROC::roc() be returned for the full dataset? 2) Should pROC::roc() be returned on each of the part datasets? Only works with get_all_aucs_fun=get_all_aucs after get_all_aucs_norecalculation() does not calculate the ROC curves for each restriction separately.

do_parallel

get_all_aucs() has parallelization enabled, but for some reason it seemed to not improve the speed of the calculation. Therefore throws an error. get_all_aucs_norecalculation() does not use it at all.

check_positive_negative_count

Pure checking/testing parameter, you would not set that TRUE anytime. Just enables checks if the number of positives/negatives was extracted correctly for the restrictions

Value

List of two elements:

"positive_label": Label of the positive class
"joined_aucs": Table with the following columns:
 threshold
     The threshold which was used as restriction value.
     "high"-part is always >= threshold
     "low"-part is always   < threshold
 auc_high
     restricted AUC for the high part, including the scaling factor
 positives_high
     How many positives are in the restricted range of high values
 negatives_high
     How many negatives are in the restricted range of high values
 scaling_high
     Scaling factor which is multiplied with the actual partial
     area under the curve to obtain the "recalculated" area under
     the curve if it was RE-calculated on the samples being in the
     restricted range of high values
 auc_var_H0_high
     Estimated variance under the nullhypothesis using
     \deqn{\frac{n_{positives} + n_{negatives} + 1}{12\cdot n_{positives} \cdot n_{negatives}}}
 rzAUC_high
     restricted standardized AUC, obtained via
     \deqn{\frac{auc_high - .5}{\sqrt{auc_var_H0_high}}}
 pval_asym_onesided_high
     Asymptotic, onesided (is AUC bigger) p-value of the restricted standardized AUC,
     obtained via:
     \deqn{1 - pnorm(full_df[["rzAUC"]])}
     Here the requirements are not fullfilled, use with utmost caution!
 pval_asym_high
     Asymptotic, twosided (is AUC different) p-value of the restricted standardized AUC,
     obtained via:
     \deqn{(1 - pnorm(abs(full_df[["rzAUC"]]))) * 2}
     Here the requirements are not fullfilled, use with utmost caution!
 auc_low
     restricted AUC for the low part, including the scaling factor
 positives_low
     How many positives are in the restricted range of low values
 negatives_low
     How many negatives are in the restricted range of low values
 scaling_low
     Scaling factor which is multiplied with the actual partial
     area under the curve to obtain the "recalculated" area under
     the curve if it was RE-calculated on the samples being in the
     restricted range of low values
 auc_var_H0_low
     Estimated variance under the nullhypothesis using
     \deqn{\frac{n_{positives} + n_{negatives} + 1}{12\cdot n_{positives} \cdot n_{negatives}}}
 rzAUC_low
     restricted standardized AUC, obtained via
     \deqn{\frac{auc_low - .5}{\sqrt{auc_var_H0_low}}}
 pval_asym_onesided_low
     Asymptotic, onesided (is AUC bigger) p-value of the restricted standardized AUC,
     obtained via:
     \deqn{1 - pnorm(full_df[["rzAUC"]])}
     Here the requirements are not fullfilled, use with utmost caution!
 pval_asym_low
     Asymptotic, twosided (is AUC different) p-value of the restricted standardized AUC,
     obtained via:
     \deqn{(1 - pnorm(abs(full_df[["rzAUC"]]))) * 2}
     Here the requirements are not fullfilled, use with utmost caution!
 tp
     Number of true positives at that threshold including all samples
 fp
     Number of false positives at that threshold including all samples
 tpr_global
     True positive rate at that threshold including all ("global") samples
 fpr_global
     False positive rate at that threshold including all ("global") samples

Examples

data(aSAH, package = "pROC")
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka
)
#> Positive label not given, setting to last level of factor: Poor
#> $joined_aucs
#> # A tibble: 110 × 21
#>    threshold auc_high positives_high negatives_high scaling_high auc_var_H0_high
#>        <dbl>    <dbl>          <dbl>          <dbl>        <dbl>           <dbl>
#>  1   -Inf       0.612             41             72         1            0.00322
#>  2      3.44    0.606             41             71         1.01         0.00323
#>  3      4.24    0.622             40             71         1.04         0.00329
#>  4      4.82    0.616             40             70         1.05         0.00330
#>  5      5.11    0.611             40             69         1.07         0.00332
#>  6      5.18    0.626             39             69         1.10         0.00338
#>  7      5.28    0.643             38             69         1.13         0.00343
#>  8      5.68    0.660             37             69         1.16         0.00349
#>  9      6.00    0.655             37             68         1.17         0.00351
#> 10      6.15    0.650             37             67         1.19         0.00353
#> # ℹ 100 more rows
#> # ℹ 15 more variables: rzAUC_high <dbl>, pval_asym_onesided_high <dbl>,
#> #   pval_asym_high <dbl>, auc_low <dbl>, positives_low <dbl>,
#> #   negatives_low <dbl>, scaling_low <dbl>, auc_var_H0_low <dbl>,
#> #   rzAUC_low <dbl>, pval_asym_onesided_low <dbl>, pval_asym_low <dbl>,
#> #   tp <dbl>, fp <dbl>, tpr_global <dbl>, fpr_global <dbl>
#> 
#> $positive_label
#> [1] "Poor"
#> 
#> attr(,"class")
#> [1] "simple_rROC" "list"       
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka,
    positive_label = "Poor"
)
#> $joined_aucs
#> # A tibble: 110 × 21
#>    threshold auc_high positives_high negatives_high scaling_high auc_var_H0_high
#>        <dbl>    <dbl>          <dbl>          <dbl>        <dbl>           <dbl>
#>  1   -Inf       0.612             41             72         1            0.00322
#>  2      3.44    0.606             41             71         1.01         0.00323
#>  3      4.24    0.622             40             71         1.04         0.00329
#>  4      4.82    0.616             40             70         1.05         0.00330
#>  5      5.11    0.611             40             69         1.07         0.00332
#>  6      5.18    0.626             39             69         1.10         0.00338
#>  7      5.28    0.643             38             69         1.13         0.00343
#>  8      5.68    0.660             37             69         1.16         0.00349
#>  9      6.00    0.655             37             68         1.17         0.00351
#> 10      6.15    0.650             37             67         1.19         0.00353
#> # ℹ 100 more rows
#> # ℹ 15 more variables: rzAUC_high <dbl>, pval_asym_onesided_high <dbl>,
#> #   pval_asym_high <dbl>, auc_low <dbl>, positives_low <dbl>,
#> #   negatives_low <dbl>, scaling_low <dbl>, auc_var_H0_low <dbl>,
#> #   rzAUC_low <dbl>, pval_asym_onesided_low <dbl>, pval_asym_low <dbl>,
#> #   tp <dbl>, fp <dbl>, tpr_global <dbl>, fpr_global <dbl>
#> 
#> $positive_label
#> [1] "Poor"
#> 
#> attr(,"class")
#> [1] "simple_rROC" "list"       
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka,
    positive_label = "Good"
)
#> $joined_aucs
#> # A tibble: 110 × 21
#>    threshold auc_high positives_high negatives_high scaling_high auc_var_H0_high
#>        <dbl>    <dbl>          <dbl>          <dbl>        <dbl>           <dbl>
#>  1   -Inf       0.388             72             41         1            0.00322
#>  2      3.44    0.394             71             41         1.01         0.00323
#>  3      4.24    0.378             71             40         1.04         0.00329
#>  4      4.82    0.384             70             40         1.05         0.00330
#>  5      5.11    0.389             69             40         1.07         0.00332
#>  6      5.18    0.374             69             39         1.10         0.00338
#>  7      5.28    0.357             69             38         1.13         0.00343
#>  8      5.68    0.340             69             37         1.16         0.00349
#>  9      6.00    0.345             68             37         1.17         0.00351
#> 10      6.15    0.350             67             37         1.19         0.00353
#> # ℹ 100 more rows
#> # ℹ 15 more variables: rzAUC_high <dbl>, pval_asym_onesided_high <dbl>,
#> #   pval_asym_high <dbl>, auc_low <dbl>, positives_low <dbl>,
#> #   negatives_low <dbl>, scaling_low <dbl>, auc_var_H0_low <dbl>,
#> #   rzAUC_low <dbl>, pval_asym_onesided_low <dbl>, pval_asym_low <dbl>,
#> #   tp <dbl>, fp <dbl>, tpr_global <dbl>, fpr_global <dbl>
#> 
#> $positive_label
#> [1] "Good"
#> 
#> attr(,"class")
#> [1] "simple_rROC" "list"       
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka,
    get_all_aucs_fun = restrictedROC:::get_all_aucs_norecalculation
)
#> Positive label not given, setting to last level of factor: Poor
#> $joined_aucs
#> # A tibble: 110 × 21
#>    threshold auc_high positives_high negatives_high scaling_high auc_var_H0_high
#>        <dbl>    <dbl>          <dbl>          <dbl>        <dbl>           <dbl>
#>  1   -Inf       0.612             41             72         1            0.00322
#>  2      3.44    0.606             41             71         1.01         0.00323
#>  3      4.24    0.622             40             71         1.04         0.00329
#>  4      4.82    0.616             40             70         1.05         0.00330
#>  5      5.11    0.611             40             69         1.07         0.00332
#>  6      5.18    0.626             39             69         1.10         0.00338
#>  7      5.28    0.643             38             69         1.13         0.00343
#>  8      5.68    0.660             37             69         1.16         0.00349
#>  9      6.00    0.655             37             68         1.17         0.00351
#> 10      6.15    0.650             37             67         1.19         0.00353
#> # ℹ 100 more rows
#> # ℹ 15 more variables: rzAUC_high <dbl>, pval_asym_onesided_high <dbl>,
#> #   pval_asym_high <dbl>, auc_low <dbl>, positives_low <dbl>,
#> #   negatives_low <dbl>, scaling_low <dbl>, auc_var_H0_low <dbl>,
#> #   rzAUC_low <dbl>, pval_asym_onesided_low <dbl>, pval_asym_low <dbl>,
#> #   tp <dbl>, fp <dbl>, tpr_global <dbl>, fpr_global <dbl>
#> 
#> $positive_label
#> [1] "Poor"
#> 
#> attr(,"class")
#> [1] "simple_rROC" "list"       
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka,
    get_all_aucs_fun = restrictedROC:::get_all_aucs
)
#> Positive label not given, setting to last level of factor: Poor
#> $joined_aucs
#> # A tibble: 110 × 19
#>    threshold auc_high positives_high negatives_high auc_var_H0_high rzAUC_high
#>        <dbl>    <dbl>          <int>          <int>           <dbl>      <dbl>
#>  1   -Inf       0.612             41             72         0.00322       1.97
#>  2      3.44    0.606             41             71         0.00323       1.87
#>  3      4.24    0.622             40             71         0.00329       2.12
#>  4      4.82    0.616             40             70         0.00330       2.02
#>  5      5.11    0.611             40             69         0.00332       1.92
#>  6      5.18    0.626             39             69         0.00338       2.17
#>  7      5.28    0.643             38             69         0.00343       2.44
#>  8      5.68    0.660             37             69         0.00349       2.71
#>  9      6.00    0.655             37             68         0.00351       2.62
#> 10      6.15    0.650             37             67         0.00353       2.53
#> # ℹ 100 more rows
#> # ℹ 13 more variables: pval_asym_onesided_high <dbl>, pval_asym_high <dbl>,
#> #   auc_low <dbl>, positives_low <int>, negatives_low <int>,
#> #   auc_var_H0_low <dbl>, rzAUC_low <dbl>, pval_asym_onesided_low <dbl>,
#> #   pval_asym_low <dbl>, tp <dbl>, fp <dbl>, tpr_global <dbl>, fpr_global <dbl>
#> 
#> $positive_label
#> [1] "Poor"
#> 
#> attr(,"class")
#> [1] "simple_rROC" "list"       
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka,
    get_all_aucs_fun = restrictedROC:::get_all_aucs,
    return_proc = TRUE
)
#> Positive label not given, setting to last level of factor: Poor
#> $joined_aucs
#> # A tibble: 110 × 19
#>    threshold auc_high positives_high negatives_high auc_var_H0_high rzAUC_high
#>        <dbl>    <dbl>          <int>          <int>           <dbl>      <dbl>
#>  1   -Inf       0.612             41             72         0.00322       1.97
#>  2      3.44    0.606             41             71         0.00323       1.87
#>  3      4.24    0.622             40             71         0.00329       2.12
#>  4      4.82    0.616             40             70         0.00330       2.02
#>  5      5.11    0.611             40             69         0.00332       1.92
#>  6      5.18    0.626             39             69         0.00338       2.17
#>  7      5.28    0.643             38             69         0.00343       2.44
#>  8      5.68    0.660             37             69         0.00349       2.71
#>  9      6.00    0.655             37             68         0.00351       2.62
#> 10      6.15    0.650             37             67         0.00353       2.53
#> # ℹ 100 more rows
#> # ℹ 13 more variables: pval_asym_onesided_high <dbl>, pval_asym_high <dbl>,
#> #   auc_low <dbl>, positives_low <int>, negatives_low <int>,
#> #   auc_var_H0_low <dbl>, rzAUC_low <dbl>, pval_asym_onesided_low <dbl>,
#> #   pval_asym_low <dbl>, tp <dbl>, fp <dbl>, tpr_global <dbl>, fpr_global <dbl>
#> 
#> $positive_label
#> [1] "Poor"
#> 
#> $pROC_lowpart
#> $pROC_lowpart$`-Inf`
#> NULL
#> 
#> $pROC_lowpart$`3.44`
#> NULL
#> 
#> $pROC_lowpart$`4.24`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 1 controls (part_df[["true"]] FALSE) < 1 cases (part_df[["true"]] TRUE).
#> Area under the curve: 1
#> 
#> $pROC_lowpart$`4.82`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 2 controls (part_df[["true"]] FALSE) < 1 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5
#> 
#> $pROC_lowpart$`5.105`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 3 controls (part_df[["true"]] FALSE) < 1 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.3333
#> 
#> $pROC_lowpart$`5.185`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 3 controls (part_df[["true"]] FALSE) < 2 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6667
#> 
#> $pROC_lowpart$`5.28`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 3 controls (part_df[["true"]] FALSE) < 3 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.7778
#> 
#> $pROC_lowpart$`5.685`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 3 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.8333
#> 
#> $pROC_lowpart$`6.005`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 4 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.625
#> 
#> $pROC_lowpart$`6.15`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 5 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5
#> 
#> $pROC_lowpart$`6.295`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 6 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4167
#> 
#> $pROC_lowpart$`6.345`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 7 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.3571
#> 
#> $pROC_lowpart$`6.465`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 8 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.3125
#> 
#> $pROC_lowpart$`6.565`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 9 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.2778
#> 
#> $pROC_lowpart$`6.69`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 10 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.25
#> 
#> $pROC_lowpart$`6.925`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 11 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.2273
#> 
#> $pROC_lowpart$`7.24`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 12 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.2083
#> 
#> $pROC_lowpart$`7.525`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 12 controls (part_df[["true"]] FALSE) < 5 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.3667
#> 
#> $pROC_lowpart$`7.645`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 13 controls (part_df[["true"]] FALSE) < 5 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.3385
#> 
#> $pROC_lowpart$`7.705`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 14 controls (part_df[["true"]] FALSE) < 5 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.3143
#> 
#> $pROC_lowpart$`7.855`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 15 controls (part_df[["true"]] FALSE) < 5 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.2933
#> 
#> $pROC_lowpart$`7.99`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 16 controls (part_df[["true"]] FALSE) < 5 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.275
#> 
#> $pROC_lowpart$`8.055`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 17 controls (part_df[["true"]] FALSE) < 5 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.2588
#> 
#> $pROC_lowpart$`8.16`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 18 controls (part_df[["true"]] FALSE) < 5 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.2444
#> 
#> $pROC_lowpart$`8.305`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 18 controls (part_df[["true"]] FALSE) < 6 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.3704
#> 
#> $pROC_lowpart$`8.455`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 18 controls (part_df[["true"]] FALSE) < 7 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4603
#> 
#> $pROC_lowpart$`8.535`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 19 controls (part_df[["true"]] FALSE) < 7 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4361
#> 
#> $pROC_lowpart$`8.72`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 20 controls (part_df[["true"]] FALSE) < 7 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4143
#> 
#> $pROC_lowpart$`8.955`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 20 controls (part_df[["true"]] FALSE) < 8 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4875
#> 
#> $pROC_lowpart$`9.225`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 21 controls (part_df[["true"]] FALSE) < 8 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4643
#> 
#> $pROC_lowpart$`9.455`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 23 controls (part_df[["true"]] FALSE) < 8 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4239
#> 
#> $pROC_lowpart$`9.52`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 24 controls (part_df[["true"]] FALSE) < 8 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4062
#> 
#> $pROC_lowpart$`9.6`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 24 controls (part_df[["true"]] FALSE) < 9 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4722
#> 
#> $pROC_lowpart$`9.665`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 25 controls (part_df[["true"]] FALSE) < 10 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.506
#> 
#> $pROC_lowpart$`9.75`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 26 controls (part_df[["true"]] FALSE) < 10 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4865
#> 
#> $pROC_lowpart$`9.805`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 27 controls (part_df[["true"]] FALSE) < 10 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4685
#> 
#> $pROC_lowpart$`9.82`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 28 controls (part_df[["true"]] FALSE) < 10 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4518
#> 
#> $pROC_lowpart$`9.84`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 29 controls (part_df[["true"]] FALSE) < 10 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4362
#> 
#> $pROC_lowpart$`9.9`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 29 controls (part_df[["true"]] FALSE) < 11 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4875
#> 
#> $pROC_lowpart$`10.14`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 30 controls (part_df[["true"]] FALSE) < 11 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4712
#> 
#> $pROC_lowpart$`10.365`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 31 controls (part_df[["true"]] FALSE) < 11 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.456
#> 
#> $pROC_lowpart$`10.41`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 31 controls (part_df[["true"]] FALSE) < 12 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5013
#> 
#> $pROC_lowpart$`10.465`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 32 controls (part_df[["true"]] FALSE) < 12 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4857
#> 
#> $pROC_lowpart$`10.53`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 33 controls (part_df[["true"]] FALSE) < 12 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.471
#> 
#> $pROC_lowpart$`10.575`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 34 controls (part_df[["true"]] FALSE) < 12 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4571
#> 
#> $pROC_lowpart$`10.715`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 35 controls (part_df[["true"]] FALSE) < 12 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.444
#> 
#> $pROC_lowpart$`10.95`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 36 controls (part_df[["true"]] FALSE) < 12 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4317
#> 
#> $pROC_lowpart$`11.08`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 37 controls (part_df[["true"]] FALSE) < 12 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.42
#> 
#> $pROC_lowpart$`11.345`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 37 controls (part_df[["true"]] FALSE) < 13 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4647
#> 
#> $pROC_lowpart$`11.635`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 37 controls (part_df[["true"]] FALSE) < 14 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5029
#> 
#> $pROC_lowpart$`11.675`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 37 controls (part_df[["true"]] FALSE) < 15 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.536
#> 
#> $pROC_lowpart$`11.7`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 38 controls (part_df[["true"]] FALSE) < 15 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5219
#> 
#> $pROC_lowpart$`11.725`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 38 controls (part_df[["true"]] FALSE) < 16 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5518
#> 
#> $pROC_lowpart$`11.85`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 39 controls (part_df[["true"]] FALSE) < 16 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5377
#> 
#> $pROC_lowpart$`12.095`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 40 controls (part_df[["true"]] FALSE) < 16 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5242
#> 
#> $pROC_lowpart$`12.375`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 40 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5522
#> 
#> $pROC_lowpart$`12.55`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 41 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5387
#> 
#> $pROC_lowpart$`12.58`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 42 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5259
#> 
#> $pROC_lowpart$`12.63`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 43 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5137
#> 
#> $pROC_lowpart$`12.69`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 44 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.502
#> 
#> $pROC_lowpart$`12.73`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 45 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4908
#> 
#> $pROC_lowpart$`12.775`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 45 controls (part_df[["true"]] FALSE) < 18 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5191
#> 
#> $pROC_lowpart$`12.85`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 46 controls (part_df[["true"]] FALSE) < 18 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5079
#> 
#> $pROC_lowpart$`12.94`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 47 controls (part_df[["true"]] FALSE) < 19 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.523
#> 
#> $pROC_lowpart$`13.05`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 48 controls (part_df[["true"]] FALSE) < 19 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5121
#> 
#> $pROC_lowpart$`13.16`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 49 controls (part_df[["true"]] FALSE) < 19 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5016
#> 
#> $pROC_lowpart$`13.305`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 49 controls (part_df[["true"]] FALSE) < 20 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5265
#> 
#> $pROC_lowpart$`13.43`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 50 controls (part_df[["true"]] FALSE) < 20 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.516
#> 
#> $pROC_lowpart$`13.505`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 51 controls (part_df[["true"]] FALSE) < 20 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5059
#> 
#> $pROC_lowpart$`13.615`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 51 controls (part_df[["true"]] FALSE) < 21 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5294
#> 
#> $pROC_lowpart$`13.77`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 51 controls (part_df[["true"]] FALSE) < 22 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5508
#> 
#> $pROC_lowpart$`13.955`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 52 controls (part_df[["true"]] FALSE) < 22 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5402
#> 
#> $pROC_lowpart$`14.15`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 52 controls (part_df[["true"]] FALSE) < 23 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5602
#> 
#> $pROC_lowpart$`14.3`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 52 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5785
#> 
#> $pROC_lowpart$`14.455`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 53 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5676
#> 
#> $pROC_lowpart$`15.055`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 54 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5571
#> 
#> $pROC_lowpart$`15.715`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 55 controls (part_df[["true"]] FALSE) < 25 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5647
#> 
#> $pROC_lowpart$`15.925`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 55 controls (part_df[["true"]] FALSE) < 26 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5815
#> 
#> $pROC_lowpart$`16.035`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 55 controls (part_df[["true"]] FALSE) < 27 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.597
#> 
#> $pROC_lowpart$`16.66`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 56 controls (part_df[["true"]] FALSE) < 27 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5863
#> 
#> $pROC_lowpart$`17.255`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 57 controls (part_df[["true"]] FALSE) < 27 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.576
#> 
#> $pROC_lowpart$`17.35`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 58 controls (part_df[["true"]] FALSE) < 27 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5661
#> 
#> $pROC_lowpart$`17.63`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 58 controls (part_df[["true"]] FALSE) < 28 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5816
#> 
#> $pROC_lowpart$`18.035`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 59 controls (part_df[["true"]] FALSE) < 28 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5717
#> 
#> $pROC_lowpart$`18.835`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 60 controls (part_df[["true"]] FALSE) < 28 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5622
#> 
#> $pROC_lowpart$`20.105`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 61 controls (part_df[["true"]] FALSE) < 28 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.553
#> 
#> $pROC_lowpart$`20.985`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 62 controls (part_df[["true"]] FALSE) < 28 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5441
#> 
#> $pROC_lowpart$`21.35`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 62 controls (part_df[["true"]] FALSE) < 29 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5598
#> 
#> $pROC_lowpart$`21.525`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 62 controls (part_df[["true"]] FALSE) < 30 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5745
#> 
#> $pROC_lowpart$`21.75`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 62 controls (part_df[["true"]] FALSE) < 31 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5882
#> 
#> $pROC_lowpart$`22.1`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 62 controls (part_df[["true"]] FALSE) < 32 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6011
#> 
#> $pROC_lowpart$`22.35`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 63 controls (part_df[["true"]] FALSE) < 32 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5915
#> 
#> $pROC_lowpart$`22.53`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 64 controls (part_df[["true"]] FALSE) < 32 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5823
#> 
#> $pROC_lowpart$`23.605`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 64 controls (part_df[["true"]] FALSE) < 33 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5949
#> 
#> $pROC_lowpart$`25.885`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 65 controls (part_df[["true"]] FALSE) < 33 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5858
#> 
#> $pROC_lowpart$`27.84`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 66 controls (part_df[["true"]] FALSE) < 33 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5769
#> 
#> $pROC_lowpart$`30.43`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 67 controls (part_df[["true"]] FALSE) < 33 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5683
#> 
#> $pROC_lowpart$`32.39`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 67 controls (part_df[["true"]] FALSE) < 34 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.581
#> 
#> $pROC_lowpart$`33.235`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 67 controls (part_df[["true"]] FALSE) < 35 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.593
#> 
#> $pROC_lowpart$`37.2`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 67 controls (part_df[["true"]] FALSE) < 36 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6043
#> 
#> $pROC_lowpart$`40.885`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 67 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.615
#> 
#> $pROC_lowpart$`44.13`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 68 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6059
#> 
#> $pROC_lowpart$`47.22`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 69 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5971
#> 
#> $pROC_lowpart$`48.775`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 69 controls (part_df[["true"]] FALSE) < 38 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6077
#> 
#> $pROC_lowpart$`52.38`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 70 controls (part_df[["true"]] FALSE) < 38 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5991
#> 
#> $pROC_lowpart$`56.825`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 70 controls (part_df[["true"]] FALSE) < 39 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6093
#> 
#> $pROC_lowpart$`65.7`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 71 controls (part_df[["true"]] FALSE) < 39 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6008
#> 
#> $pROC_lowpart$`76.435`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 71 controls (part_df[["true"]] FALSE) < 40 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6107
#> 
#> $pROC_lowpart$`249.745`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 72 controls (part_df[["true"]] FALSE) < 40 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6023
#> 
#> $pROC_lowpart$`Inf`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 72 controls (part_df[["true"]] FALSE) < 41 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.612
#> 
#> 
#> $pROC_highpart
#> $pROC_highpart$`-Inf`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 72 controls (part_df[["true"]] FALSE) < 41 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.612
#> 
#> $pROC_highpart$`3.44`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 71 controls (part_df[["true"]] FALSE) < 41 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6065
#> 
#> $pROC_highpart$`4.24`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 71 controls (part_df[["true"]] FALSE) < 40 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6217
#> 
#> $pROC_highpart$`4.82`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 70 controls (part_df[["true"]] FALSE) < 40 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6162
#> 
#> $pROC_highpart$`5.105`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 69 controls (part_df[["true"]] FALSE) < 40 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6107
#> 
#> $pROC_highpart$`5.185`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 69 controls (part_df[["true"]] FALSE) < 39 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6263
#> 
#> $pROC_highpart$`5.28`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 69 controls (part_df[["true"]] FALSE) < 38 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6428
#> 
#> $pROC_highpart$`5.685`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 69 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6602
#> 
#> $pROC_highpart$`6.005`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 68 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6552
#> 
#> $pROC_highpart$`6.15`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 67 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6501
#> 
#> $pROC_highpart$`6.295`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 66 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6448
#> 
#> $pROC_highpart$`6.345`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 65 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6393
#> 
#> $pROC_highpart$`6.465`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 64 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6337
#> 
#> $pROC_highpart$`6.565`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 63 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6278
#> 
#> $pROC_highpart$`6.69`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 62 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6218
#> 
#> $pROC_highpart$`6.925`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 61 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6156
#> 
#> $pROC_highpart$`7.24`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 60 controls (part_df[["true"]] FALSE) < 37 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6092
#> 
#> $pROC_highpart$`7.525`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 60 controls (part_df[["true"]] FALSE) < 36 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6262
#> 
#> $pROC_highpart$`7.645`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 59 controls (part_df[["true"]] FALSE) < 36 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6198
#> 
#> $pROC_highpart$`7.705`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 58 controls (part_df[["true"]] FALSE) < 36 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6133
#> 
#> $pROC_highpart$`7.855`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 57 controls (part_df[["true"]] FALSE) < 36 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6065
#> 
#> $pROC_highpart$`7.99`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 56 controls (part_df[["true"]] FALSE) < 36 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5995
#> 
#> $pROC_highpart$`8.055`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 55 controls (part_df[["true"]] FALSE) < 36 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5922
#> 
#> $pROC_highpart$`8.16`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 54 controls (part_df[["true"]] FALSE) < 36 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5846
#> 
#> $pROC_highpart$`8.305`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 54 controls (part_df[["true"]] FALSE) < 35 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6013
#> 
#> $pROC_highpart$`8.455`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 54 controls (part_df[["true"]] FALSE) < 34 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.619
#> 
#> $pROC_highpart$`8.535`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 53 controls (part_df[["true"]] FALSE) < 34 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6118
#> 
#> $pROC_highpart$`8.72`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 52 controls (part_df[["true"]] FALSE) < 34 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6044
#> 
#> $pROC_highpart$`8.955`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 52 controls (part_df[["true"]] FALSE) < 33 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6227
#> 
#> $pROC_highpart$`9.225`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 51 controls (part_df[["true"]] FALSE) < 33 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6153
#> 
#> $pROC_highpart$`9.455`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 49 controls (part_df[["true"]] FALSE) < 33 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5996
#> 
#> $pROC_highpart$`9.52`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 48 controls (part_df[["true"]] FALSE) < 33 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5912
#> 
#> $pROC_highpart$`9.6`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 48 controls (part_df[["true"]] FALSE) < 32 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6097
#> 
#> $pROC_highpart$`9.665`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 47 controls (part_df[["true"]] FALSE) < 31 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6211
#> 
#> $pROC_highpart$`9.75`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 46 controls (part_df[["true"]] FALSE) < 31 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6129
#> 
#> $pROC_highpart$`9.805`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 45 controls (part_df[["true"]] FALSE) < 31 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6043
#> 
#> $pROC_highpart$`9.82`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 44 controls (part_df[["true"]] FALSE) < 31 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5953
#> 
#> $pROC_highpart$`9.84`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 43 controls (part_df[["true"]] FALSE) < 31 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5859
#> 
#> $pROC_highpart$`9.9`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 43 controls (part_df[["true"]] FALSE) < 30 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6054
#> 
#> $pROC_highpart$`10.14`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 42 controls (part_df[["true"]] FALSE) < 30 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.596
#> 
#> $pROC_highpart$`10.365`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 41 controls (part_df[["true"]] FALSE) < 30 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5862
#> 
#> $pROC_highpart$`10.41`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 41 controls (part_df[["true"]] FALSE) < 29 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6064
#> 
#> $pROC_highpart$`10.465`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 40 controls (part_df[["true"]] FALSE) < 29 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5966
#> 
#> $pROC_highpart$`10.53`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 39 controls (part_df[["true"]] FALSE) < 29 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5862
#> 
#> $pROC_highpart$`10.575`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 38 controls (part_df[["true"]] FALSE) < 29 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5753
#> 
#> $pROC_highpart$`10.715`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 37 controls (part_df[["true"]] FALSE) < 29 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5638
#> 
#> $pROC_highpart$`10.95`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 36 controls (part_df[["true"]] FALSE) < 29 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5517
#> 
#> $pROC_highpart$`11.08`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 35 controls (part_df[["true"]] FALSE) < 29 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5389
#> 
#> $pROC_highpart$`11.345`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 35 controls (part_df[["true"]] FALSE) < 28 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5582
#> 
#> $pROC_highpart$`11.635`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 35 controls (part_df[["true"]] FALSE) < 27 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5788
#> 
#> $pROC_highpart$`11.675`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 35 controls (part_df[["true"]] FALSE) < 26 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6011
#> 
#> $pROC_highpart$`11.7`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 34 controls (part_df[["true"]] FALSE) < 26 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5894
#> 
#> $pROC_highpart$`11.725`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 34 controls (part_df[["true"]] FALSE) < 25 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6129
#> 
#> $pROC_highpart$`11.85`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 33 controls (part_df[["true"]] FALSE) < 25 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6012
#> 
#> $pROC_highpart$`12.095`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 32 controls (part_df[["true"]] FALSE) < 25 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5888
#> 
#> $pROC_highpart$`12.375`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 32 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6133
#> 
#> $pROC_highpart$`12.55`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 31 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6008
#> 
#> $pROC_highpart$`12.58`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 30 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5875
#> 
#> $pROC_highpart$`12.63`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 29 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5733
#> 
#> $pROC_highpart$`12.69`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 28 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.558
#> 
#> $pROC_highpart$`12.73`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 27 controls (part_df[["true"]] FALSE) < 24 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5417
#> 
#> $pROC_highpart$`12.775`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 27 controls (part_df[["true"]] FALSE) < 23 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5652
#> 
#> $pROC_highpart$`12.85`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 26 controls (part_df[["true"]] FALSE) < 23 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5485
#> 
#> $pROC_highpart$`12.94`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 25 controls (part_df[["true"]] FALSE) < 22 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5555
#> 
#> $pROC_highpart$`13.05`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 24 controls (part_df[["true"]] FALSE) < 22 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5369
#> 
#> $pROC_highpart$`13.16`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 23 controls (part_df[["true"]] FALSE) < 22 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5168
#> 
#> $pROC_highpart$`13.305`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 23 controls (part_df[["true"]] FALSE) < 21 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5414
#> 
#> $pROC_highpart$`13.43`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 22 controls (part_df[["true"]] FALSE) < 21 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5206
#> 
#> $pROC_highpart$`13.505`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 21 controls (part_df[["true"]] FALSE) < 21 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4977
#> 
#> $pROC_highpart$`13.615`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 21 controls (part_df[["true"]] FALSE) < 20 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5226
#> 
#> $pROC_highpart$`13.77`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 21 controls (part_df[["true"]] FALSE) < 19 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5501
#> 
#> $pROC_highpart$`13.955`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 20 controls (part_df[["true"]] FALSE) < 19 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5276
#> 
#> $pROC_highpart$`14.15`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 20 controls (part_df[["true"]] FALSE) < 18 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5569
#> 
#> $pROC_highpart$`14.3`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 20 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5897
#> 
#> $pROC_highpart$`14.455`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 19 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5681
#> 
#> $pROC_highpart$`15.055`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 18 controls (part_df[["true"]] FALSE) < 17 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5441
#> 
#> $pROC_highpart$`15.715`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 17 controls (part_df[["true"]] FALSE) < 16 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5515
#> 
#> $pROC_highpart$`15.925`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 17 controls (part_df[["true"]] FALSE) < 15 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5882
#> 
#> $pROC_highpart$`16.035`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 17 controls (part_df[["true"]] FALSE) < 14 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6303
#> 
#> $pROC_highpart$`16.66`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 16 controls (part_df[["true"]] FALSE) < 14 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6071
#> 
#> $pROC_highpart$`17.255`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 15 controls (part_df[["true"]] FALSE) < 14 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.581
#> 
#> $pROC_highpart$`17.35`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 14 controls (part_df[["true"]] FALSE) < 14 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.551
#> 
#> $pROC_highpart$`17.63`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 14 controls (part_df[["true"]] FALSE) < 13 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5934
#> 
#> $pROC_highpart$`18.035`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 13 controls (part_df[["true"]] FALSE) < 13 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5621
#> 
#> $pROC_highpart$`18.835`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 12 controls (part_df[["true"]] FALSE) < 13 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5256
#> 
#> $pROC_highpart$`20.105`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 11 controls (part_df[["true"]] FALSE) < 13 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4825
#> 
#> $pROC_highpart$`20.985`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 10 controls (part_df[["true"]] FALSE) < 13 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4308
#> 
#> $pROC_highpart$`21.35`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 10 controls (part_df[["true"]] FALSE) < 12 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4667
#> 
#> $pROC_highpart$`21.525`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 10 controls (part_df[["true"]] FALSE) < 11 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5091
#> 
#> $pROC_highpart$`21.75`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 10 controls (part_df[["true"]] FALSE) < 10 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.56
#> 
#> $pROC_highpart$`22.1`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 10 controls (part_df[["true"]] FALSE) < 9 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6222
#> 
#> $pROC_highpart$`22.35`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 9 controls (part_df[["true"]] FALSE) < 9 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5802
#> 
#> $pROC_highpart$`22.53`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 8 controls (part_df[["true"]] FALSE) < 9 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5278
#> 
#> $pROC_highpart$`23.605`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 8 controls (part_df[["true"]] FALSE) < 8 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5938
#> 
#> $pROC_highpart$`25.885`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 7 controls (part_df[["true"]] FALSE) < 8 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5357
#> 
#> $pROC_highpart$`27.84`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 6 controls (part_df[["true"]] FALSE) < 8 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4583
#> 
#> $pROC_highpart$`30.43`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 5 controls (part_df[["true"]] FALSE) < 8 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.35
#> 
#> $pROC_highpart$`32.39`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 5 controls (part_df[["true"]] FALSE) < 7 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4
#> 
#> $pROC_highpart$`33.235`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 5 controls (part_df[["true"]] FALSE) < 6 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.4667
#> 
#> $pROC_highpart$`37.2`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 5 controls (part_df[["true"]] FALSE) < 5 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.56
#> 
#> $pROC_highpart$`40.885`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 5 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.7
#> 
#> $pROC_highpart$`44.13`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 4 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.625
#> 
#> $pROC_highpart$`47.22`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 3 controls (part_df[["true"]] FALSE) < 4 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5
#> 
#> $pROC_highpart$`48.775`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 3 controls (part_df[["true"]] FALSE) < 3 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.6667
#> 
#> $pROC_highpart$`52.38`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 2 controls (part_df[["true"]] FALSE) < 3 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5
#> 
#> $pROC_highpart$`56.825`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 2 controls (part_df[["true"]] FALSE) < 2 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.75
#> 
#> $pROC_highpart$`65.7`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 1 controls (part_df[["true"]] FALSE) < 2 cases (part_df[["true"]] TRUE).
#> Area under the curve: 0.5
#> 
#> $pROC_highpart$`76.435`
#> 
#> Call:
#> roc.default(response = part_df[["true"]], predictor = part_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: part_df[["pred"]] in 1 controls (part_df[["true"]] FALSE) < 1 cases (part_df[["true"]] TRUE).
#> Area under the curve: 1
#> 
#> $pROC_highpart$`249.745`
#> NULL
#> 
#> $pROC_highpart$`Inf`
#> NULL
#> 
#> 
#> $pROC_full
#> 
#> Call:
#> roc.default(response = true_pred_df[["true"]], predictor = true_pred_df[["pred"]],     levels = c(FALSE, TRUE), direction = direction)
#> 
#> Data: true_pred_df[["pred"]] in 72 controls (true_pred_df[["true"]] FALSE) < 41 cases (true_pred_df[["true"]] TRUE).
#> Area under the curve: 0.612
#> 
#> attr(,"class")
#> [1] "simple_rROC" "list"       
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka,
    check_positive_negative_count = TRUE
)
#> Positive label not given, setting to last level of factor: Poor
#> $joined_aucs
#> # A tibble: 110 × 21
#>    threshold auc_high positives_high negatives_high scaling_high auc_var_H0_high
#>        <dbl>    <dbl>          <dbl>          <dbl>        <dbl>           <dbl>
#>  1   -Inf       0.612             41             72         1            0.00322
#>  2      3.44    0.606             41             71         1.01         0.00323
#>  3      4.24    0.622             40             71         1.04         0.00329
#>  4      4.82    0.616             40             70         1.05         0.00330
#>  5      5.11    0.611             40             69         1.07         0.00332
#>  6      5.18    0.626             39             69         1.10         0.00338
#>  7      5.28    0.643             38             69         1.13         0.00343
#>  8      5.68    0.660             37             69         1.16         0.00349
#>  9      6.00    0.655             37             68         1.17         0.00351
#> 10      6.15    0.650             37             67         1.19         0.00353
#> # ℹ 100 more rows
#> # ℹ 15 more variables: rzAUC_high <dbl>, pval_asym_onesided_high <dbl>,
#> #   pval_asym_high <dbl>, auc_low <dbl>, positives_low <dbl>,
#> #   negatives_low <dbl>, scaling_low <dbl>, auc_var_H0_low <dbl>,
#> #   rzAUC_low <dbl>, pval_asym_onesided_low <dbl>, pval_asym_low <dbl>,
#> #   tp <dbl>, fp <dbl>, tpr_global <dbl>, fpr_global <dbl>
#> 
#> $positive_label
#> [1] "Poor"
#> 
#> attr(,"class")
#> [1] "simple_rROC" "list"       
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka,
    do_parallel = TRUE
)
#> Positive label not given, setting to last level of factor: Poor
#> $joined_aucs
#> # A tibble: 110 × 21
#>    threshold auc_high positives_high negatives_high scaling_high auc_var_H0_high
#>        <dbl>    <dbl>          <dbl>          <dbl>        <dbl>           <dbl>
#>  1   -Inf       0.612             41             72         1            0.00322
#>  2      3.44    0.606             41             71         1.01         0.00323
#>  3      4.24    0.622             40             71         1.04         0.00329
#>  4      4.82    0.616             40             70         1.05         0.00330
#>  5      5.11    0.611             40             69         1.07         0.00332
#>  6      5.18    0.626             39             69         1.10         0.00338
#>  7      5.28    0.643             38             69         1.13         0.00343
#>  8      5.68    0.660             37             69         1.16         0.00349
#>  9      6.00    0.655             37             68         1.17         0.00351
#> 10      6.15    0.650             37             67         1.19         0.00353
#> # ℹ 100 more rows
#> # ℹ 15 more variables: rzAUC_high <dbl>, pval_asym_onesided_high <dbl>,
#> #   pval_asym_high <dbl>, auc_low <dbl>, positives_low <dbl>,
#> #   negatives_low <dbl>, scaling_low <dbl>, auc_var_H0_low <dbl>,
#> #   rzAUC_low <dbl>, pval_asym_onesided_low <dbl>, pval_asym_low <dbl>,
#> #   tp <dbl>, fp <dbl>, tpr_global <dbl>, fpr_global <dbl>
#> 
#> $positive_label
#> [1] "Poor"
#> 
#> attr(,"class")
#> [1] "simple_rROC" "list"       
if (FALSE) { # \dontrun{
simple_rROC(
    response = aSAH$outcome,
    predictor = aSAH$ndka,
    get_all_aucs_fun = restrictedROC:::get_all_aucs,
    do_parallel = TRUE
)
} # }