Question

我在哪里错了？

我试图使用roc在同一个地块上绘制两条ggplot曲线。

运行以下代码时出现此错误

Error: Don't know how to add o to a plot

ggplot不起作用

ggroc(roc_1) +
  ggroc(roc_2) +
  labs(title = "ROC curve", y = "Sensitivity", x = "Specificity")

基础包工作

plot(roc_1, col = 1, lty = 2, main = "ROC")
plot(roc_2, col = 4, lty = 3, add = TRUE)

dput太大而无法在stackoverflow上发布，所以这里是ROC计算之一的结构。

我试图绘制两条类似于下面的ROC曲线。

List of 15
 $ percent           : logi FALSE
 $ sensitivities     : num [1:26455] 1 1 1 1 1 1 1 1 1 1 ...
 $ specificities     : num [1:26455] 0.00 4.00e-05 8.01e-05 1.20e-04 1.60e-04 ...
 $ thresholds        : num [1:26455] -Inf 0.0017 0.00189 0.00201 0.00214 ...
 $ direction         : chr "<"
 $ cases             : num [1:1540] 0.958 0.919 0.973 0.785 0.933 ...
 $ controls          : num [1:24975] 0.6604 0.026 0.1389 0.0558 0.0594 ...
 $ fun.sesp          :function (thresholds, controls, cases, direction)  
 $ auc               :Classes 'auc', 'numeric'  atomic [1:1] 0.942
  .. ..- attr(*, "partial.auc")= logi FALSE
  .. ..- attr(*, "percent")= logi FALSE
  .. ..- attr(*, "roc")=List of 15
  .. .. ..$ percent           : logi FALSE
  .. .. ..$ sensitivities     : num [1:26455] 1 1 1 1 1 1 1 1 1 1 ...
  .. .. ..$ specificities     : num [1:26455] 0.00 4.00e-05 8.01e-05 1.20e-04 1.60e-04 ...
  .. .. ..$ thresholds        : num [1:26455] -Inf 0.0017 0.00189 0.00201 0.00214 ...
  .. .. ..$ direction         : chr "<"
  .. .. ..$ cases             : num [1:1540] 0.958 0.919 0.973 0.785 0.933 ...
  .. .. ..$ controls          : num [1:24975] 0.6604 0.026 0.1389 0.0558 0.0594 ...
  .. .. ..$ fun.sesp          :function (thresholds, controls, cases, direction)  
  .. .. ..$ auc               :Classes 'auc', 'numeric'  atomic [1:1] 0.942
  .. .. .. .. ..- attr(*, "partial.auc")= logi FALSE
  .. .. .. .. ..- attr(*, "percent")= logi FALSE
  .. .. .. .. ..- attr(*, "roc")=List of 8
  .. .. .. .. .. ..$ percent      : logi FALSE
  .. .. .. .. .. ..$ sensitivities: num [1:26455] 1 1 1 1 1 1 1 1 1 1 ...
  .. .. .. .. .. ..$ specificities: num [1:26455] 0.00 4.00e-05 8.01e-05 1.20e-04 1.60e-04 ...
  .. .. .. .. .. ..$ thresholds   : num [1:26455] -Inf 0.0017 0.00189 0.00201 0.00214 ...
  .. .. .. .. .. ..$ direction    : chr "<"
  .. .. .. .. .. ..$ cases        : num [1:1540] 0.958 0.919 0.973 0.785 0.933 ...
  .. .. .. .. .. ..$ controls     : num [1:24975] 0.6604 0.026 0.1389 0.0558 0.0594 ...
  .. .. .. .. .. ..$ fun.sesp     :function (thresholds, controls, cases, direction)  
  .. .. .. .. .. ..- attr(*, "class")= chr "roc"
  .. .. ..$ call              : language roc.default(response = results$testactual, predictor = results$pred)
  .. .. ..$ original.predictor: num [1:26515] 0.6604 0.026 0.1389 0.0558 0.0594 ...
  .. .. ..$ original.response : int [1:26515] 0 0 0 0 0 0 0 0 0 0 ...
  .. .. ..$ predictor         : num [1:26515] 0.6604 0.026 0.1389 0.0558 0.0594 ...
  .. .. ..$ response          : int [1:26515] 0 0 0 0 0 0 0 0 0 0 ...
  .. .. ..$ levels            : chr [1:2] "0" "1"
  .. .. ..- attr(*, "class")= chr "roc"
 $ call              : language roc.default(response = results$testactual, predictor = results$pred)
 $ original.predictor: num [1:26515] 0.6604 0.026 0.1389 0.0558 0.0594 ...
 $ original.response : int [1:26515] 0 0 0 0 0 0 0 0 0 0 ...
 $ predictor         : num [1:26515] 0.6604 0.026 0.1389 0.0558 0.0594 ...
 $ response          : int [1:26515] 0 0 0 0 0 0 0 0 0 0 ...
 $ levels            : chr [1:2] "0" "1"
 - attr(*, "class")= chr "roc"

以下是基本包的外观，只是尝试在ggplot中执行此操作

编辑：临时解决问题的方法。

rocy <- cbind(roc_1$sensitivities, roc_1$specificities, roc_2$sensitivities, roc_2$specificities)
rocy <- as.data.frame(rocy)
ggplot(rocy) + 
  geom_line(aes(y = V1, x = V2)) +
  geom_line(aes(y = V3, x = V4))

编辑：解决方案

library(pROC)
rocobj1 <- roc(df$actualoutcome1, data$prediction1)
rocobj2 <- roc(df$actualoutcome1, data$prediction2)
ggroc(list(call_roc_name_1 = rocobj1, call_roc_name_2 = rocobj2))

使用roc（）在同一图上的多个roc曲线

0 个答案: