Hướng dẫn smooth roc curve python - con trăn đường cong roc mịn

Bạn đang sử dụng các dự đoán ngưỡng để tạo đường cong ROC. Thay vào đó, bạn nên sử dụng các giá trị tin cậy ban đầu, nếu không bạn sẽ chỉ nhận được 1 điểm trung gian trên đường cong.

Dưới đây là một số dữ liệu ví dụ và đường cong ROC bạn sẽ nhận được.

y_test: [1, 0, 0, 0, 0, 1, 0, 0, 0, 0, ...
y_predicted: [1, 0.405, 0.601, 0.579, 0.03, 0.98, 0.06, 0.242, 0.379, 0.09, ...
y_predicted_thresholded: [1, 0, 1, 1, 0, 1, 0, 0, 0, 0, ...

Tôi biết câu hỏi là hai năm tuổi và câu trả lời kỹ thuật đã được đưa ra trong các bình luận, nhưng một câu trả lời công phu hơn có thể giúp những người khác vẫn phải vật lộn với các khái niệm.

Đường cong ROC của OP sai vì anh ta đã sử dụng các giá trị dự đoán của các mô hình của mình thay vì xác suất.

Điều đó có nghĩa là gì?

Khi một mô hình được đào tạo, nó sẽ học được các mối quan hệ giữa các biến đầu vào và biến đầu ra. Đối với mỗi quan sát, mô hình được hiển thị, mô hình tìm hiểu mức độ có thể là một quan sát nhất định thuộc về một lớp nhất định. Khi mô hình được trình bày với dữ liệu thử nghiệm, nó sẽ đoán cho mỗi quan sát chưa từng thấy nó có thể thuộc về một lớp nhất định.

Làm thế nào để mô hình biết nếu một quan sát thuộc về một lớp? Trong quá trình thử nghiệm, mô hình nhận được một quan sát mà nó ước tính xác suất 51% thuộc về lớp X. Làm thế nào để đưa ra quyết định gắn nhãn là thuộc lớp X hay không? Nhà nghiên cứu sẽ đặt một ngưỡng nói với mô hình rằng tất cả các quan sát với xác suất dưới 50% phải được phân loại là y và tất cả những người trên phải được phân loại là X. Đôi khi nhà nghiên cứu muốn đặt ra một quy tắc chặt chẽ hơn vì họ quan tâm hơn đến chính xác Dự đoán một lớp nhất định như X thay vì cố gắng dự đoán tất cả chúng là tốt. During testing the model receives an observation for which it estimates a probability of 51% of belonging to Class X. How does take the decision to label as belonging to Class X or not? The researcher will set a threshold telling the model that all observations with a probability under 50% must be classified as Y and all those above must be classified as X. Sometimes the researcher wants to set a stricter rule because they're more interested in correctly predicting a given class like X rather than trying to predict all of them as well.

Vì vậy, mô hình được đào tạo của bạn đã ước tính xác suất cho từng quan sát của bạn, nhưng cuối cùng ngưỡng sẽ quyết định trong lớp quan sát của bạn sẽ được phân loại.

Vì sao vấn đề này?

Đường cong được tạo ra bởi ROC biểu thị một điểm cho từng tỷ lệ dương thực sự và tỷ lệ dương tính giả của mô hình của bạn ở các mức ngưỡng khác nhau. Điều này giúp nhà nghiên cứu thấy sự đánh đổi giữa FPR và TPR cho tất cả các cấp ngưỡng.

Vì vậy, khi bạn vượt qua các giá trị dự đoán thay vì xác suất dự đoán cho ROC của bạn, bạn sẽ chỉ có một điểm vì các giá trị này được tính toán bằng một ngưỡng cụ thể. Bởi vì điểm đó là TPR và FPR của mô hình của bạn cho một mức ngưỡng cụ thể.

Những gì bạn cần làm là sử dụng xác suất thay thế và để ngưỡng khác nhau.

Chạy mô hình của bạn như vậy:

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Khi tạo ma trận nhầm lẫn của bạn, bạn sẽ sử dụng các giá trị của mô hình của bạn

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

Khi tạo đường cong ROC của bạn, bạn sẽ sử dụng xác suất

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

Smooth: Smooth một đường cong ROC: Smooth a ROC curve

Mô tả đối số sử dụng đối số Chi tiết Lỗi giá trị Tài liệu tham khảo Xem thêm ví dụ

Xem Nguồn: R/Smooth.R

Hàm này làm mịn một đường cong ROC của dự đoán số. Theo mặc định, việc làm mịn nhị phân được thực hiện, nhưng mật độ hoặc làm mịn tùy chỉnh được hỗ trợ.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

smooth(...)
## Default S3 method:
smooth(...)
## S3 method for class 'roc'
smooth(roc,
method=c("binormal", "density", "fitdistr", "logcondens",
"logcondens.smooth"), n=512, bw = "nrd0", density=NULL,
density.controls=density, density.cases=density,
start=NULL, start.controls=start, start.cases=start, 
reuse.auc=TRUE, reuse.ci=FALSE, ...)
## S3 method for class 'smooth.roc'
smooth(smooth.roc, ...)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

smooth(...)
## Default S3 method:
smooth(...)
## S3 method for class 'roc'
smooth(roc,
method=c("binormal", "density", "fitdistr", "logcondens",
"logcondens.smooth"), n=512, bw = "nrd0", density=NULL,
density.controls=density, density.cases=density,
start=NULL, start.controls=start, start.cases=start, 
reuse.auc=TRUE, reuse.ci=FALSE, ...)
## S3 method for class 'smooth.roc'
smooth(smooth.roc, ...)

smooth(...)
## Default S3 method:
smooth(...)
## S3 method for class 'roc'
smooth(roc,
method=c("binormal", "density", "fitdistr", "logcondens",
"logcondens.smooth"), n=512, bw = "nrd0", density=NULL,
density.controls=density, density.cases=density,
start=NULL, start.controls=start, start.cases=start, 
reuse.auc=TRUE, reuse.ci=FALSE, ...)
## S3 method for class 'smooth.roc'
smooth(smooth.roc, ...)

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

Nếu

smooth(...)
## Default S3 method:
smooth(...)
## S3 method for class 'roc'
smooth(roc,
method=c("binormal", "density", "fitdistr", "logcondens",
"logcondens.smooth"), n=512, bw = "nrd0", density=NULL,
density.controls=density, density.cases=density,
start=NULL, start.controls=start, start.cases=start, 
reuse.auc=TRUE, reuse.ci=FALSE, ...)
## S3 method for class 'smooth.roc'
smooth(smooth.roc, ...)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Với

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

smooth(...)
## Default S3 method:
smooth(...)
## S3 method for class 'roc'
smooth(roc,
method=c("binormal", "density", "fitdistr", "logcondens",
"logcondens.smooth"), n=512, bw = "nrd0", density=NULL,
density.controls=density, density.cases=density,
start=NULL, start.controls=start, start.cases=start, 
reuse.auc=TRUE, reuse.ci=FALSE, ...)
## S3 method for class 'smooth.roc'
smooth(smooth.roc, ...)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Ngoài ra,

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Điều quan trọng là phải tôn vinh

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Nếu

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Các đường cong ROC được làm mịn có thể được truyền lại để làm mịn trở lại. Trong trường hợp này, việc làm mịn không được áp dụng lại trên đường cong ROC được làm mịn nhưng đối tượng ROC ROC ban đầu sẽ được sử dụng lại.

Lưu ý rằng một đường cong

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

Một danh sách các lớp học mịn.

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

import scikitplot as skplt
plot = skplt.metrics.plot_roc(y_test, knn_y_proba)
plt.title("ROC Curves - K-Nearest Neighbors")

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

Type 'citation("pROC")' for a citation.

Attaching package: ‘pROC’

The following objects are masked from ‘package:stats’:

    cov, smooth, var

Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
smooth.roc(roc = rocobj)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Setting levels: control = Good, case = Poor
Setting direction: controls < cases

Call:
roc.default(response = aSAH$outcome, predictor = aSAH$s100b,     smooth = TRUE)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: binormal 
Area under the curve: 0.74
Loading required namespace: MASS
Loading required namespace: logcondens

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density = dnorm,     start = list(mean = 1), sd = 0.1)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.9474
Warning messages:
1: In stats::optim(x = c(0.13, 0.14, 0.1, 0.04, 0.47, 0.18, 0.1, 0.1,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly
2: In stats::optim(x = c(0.13, 0.1, 0.16, 0.12, 0.44, 0.71, 0.49, 0.07,  :
  one-dimensional optimization by Nelder-Mead is unreliable:
use "Brent" or optimize() directly

Call:
smooth.roc(roc = rocobj, method = "fitdistr", density.controls = "normal",     density.cases = "lognormal")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: fitdistr 
Area under the curve: 0.6546

Call:
smooth.roc(roc = rocobj, method = "density", density.controls = density.controls$y,     density.cases = density.cases$y)

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244

Call:
smooth.roc(roc = rocobj, method = "density")

Data: aSAH$s100b in 72 controls (aSAH$outcome Good) < 41 cases (aSAH$outcome Poor).
Smoothing: density (bandwidth: nrd0; adjust: 1)
Area under the curve: 0.7244
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
Setting direction: controls < cases
95% CI: 0.874-0.8774 (100 stratified bootstrap replicates)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

smooth(...)
## Default S3 method:
smooth(...)
## S3 method for class 'roc'
smooth(roc,
method=c("binormal", "density", "fitdistr", "logcondens",
"logcondens.smooth"), n=512, bw = "nrd0", density=NULL,
density.controls=density, density.cases=density,
start=NULL, start.controls=start, start.cases=start, 
reuse.auc=TRUE, reuse.ci=FALSE, ...)
## S3 method for class 'smooth.roc'
smooth(smooth.roc, ...)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Thuộc tính

Ngoài ra, đối tượng

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Thông báo Chức năng 'mật độ' phải trả về một vectơ số hoặc một danh sách có mục 'y'. sẽ được hiển thị nếu hàm

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Làm mịn nhị phân không thể làm mịn đường cong ROC được xác định chỉ bằng một điểm. Bất kỳ nỗ lực nào như vậy sẽ thất bại với lỗi đường cong ROC không thể vượt qua (không đủ điểm).

Nếu đường cong ROC mịn được tạo bởi

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Phương pháp làm mịn

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Các phương thức

from mlxtend.plotting import plot_confusion_matrix
fig, ax = plot_confusion_matrix(conf_mat=confusion_matrix(y_test,knn_y_model),
                                show_absolute=True,show_normed=True,colorbar=True)
plt.title("Confusion matrix - KNN")
plt.ylabel('True label')
plt.xlabel('Predicted label'

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

James E. Hanley (1988) Sự mạnh mẽ của các giả định của Bin Binormal được sử dụng trong các đường cong ROC phù hợp. Ra quyết định y tế 8, 197 Từ203.8, 197–203.

Lutz Duembgen, Kaspar Rufibach (2011) Logcondens: Các tính toán liên quan đến ước tính mật độ log-concave đơn biến. Tạp chí phần mềm thống kê, 39, 1 trận28. URL: jstatsoft.org/v39/i06.39, 1–28. URL: jstatsoft.org/v39/i06.

Xavier Robin, Natacha Turck, Alexandre Hainard, et al. . BMC Bioinformatics, 7, 77. doi: doi: 10.1186/1471-2105-12-77.7, 77. DOI: doi: 10.1186/1471-2105-12-77.

Kaspar Rufibach (2011) Một công cụ ước tính đường cong ROC trơn tru dựa trên ước tính mật độ log-concave. Tạp chí quốc tế về sinh học, 8, được chấp nhận. Doi: doi: 10.1515/1557-4679.1378. ARXIV: 1103.1787.8, accepted. DOI: doi: 10.1515/1557-4679.1378. arXiv: 1103.1787.

William N. Venables, Brian D. Ripley (2002). Thống kê ứng dụng hiện đại với Siêu. New York, Springer. Sách Google.

Kelly H. Zou, W. J. Hall và David E. Shapiro (1997) Các đường cong đặc tính hoạt động của máy thu không tham số mịn (ROC) cho các xét nghiệm chẩn đoán liên tục. Thống kê trong Y học 18, 2143 Từ2156. Doi: doi: 10.1002/(sici) 1097-0258 (19971015) 16: 193.0.co; 2-3.18, 2143–2156. DOI: doi: 10.1002/(SICI)1097-0258(19971015)16:19<2143::AID-SIM655>3.0.CO;2-3.

from sklearn.neighbors import KNeighborsClassifier
knn = KNeighborsClassifier()
knn_model = knn.fit(X_train,y_train)
#Use the values for your confusion matrix
knn_y_model = knn_model.predict(X=X_test)
# Use the probabilities for your ROC and Precision-recall curves
knn_y_proba = knn_model.predict_proba(X=X_test)

Các gói cran Mass và logcondens được sử dụng trong chức năng này.MASS and logcondens employed in this function.

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117

data(aSAH)

##  Basic example
rocobj <- roc(aSAH$outcome, aSAH$s100b)
smooth(rocobj)
# or directly with roc()
roc(aSAH$outcome, aSAH$s100b, smooth=TRUE)

# plotting
plot(rocobj)
rs <- smooth(rocobj, method="binormal")
plot(rs, add=TRUE, col="green")
rs2 <- smooth(rocobj, method="density")
plot(rs2, add=TRUE, col="blue")
rs3 <- smooth(rocobj, method="fitdistr", density="lognormal")
plot(rs3, add=TRUE, col="magenta")
if (requireNamespace("logcondens")) {
rs4 <- smooth(rocobj, method="logcondens")
plot(rs4, add=TRUE, col="brown")
rs5 <- smooth(rocobj, method="logcondens.smooth")
plot(rs5, add=TRUE, col="orange")
}
legend("bottomright", legend=c("Empirical", "Binormal", "Density", "Log-normal",
                               "Log-concave density", "Smoothed log-concave density"),
       col=c("black", "green", "blue", "magenta", "brown", "orange"), lwd=2)

## Advanced smoothing

# if we know the distributions are normal with sd=0.1 and an unknown mean:
smooth(rocobj, method="fitdistr", density=dnorm, start=list(mean=1), sd=.1)
# different distibutions for controls and cases:
smooth(rocobj, method="fitdistr", density.controls="normal", density.cases="lognormal")

# with densities
bw <- bw.nrd0(rocobj$predictor)
density.controls <- density(rocobj$controls, from=min(rocobj$predictor) - 3 * bw,
                            to=max(rocobj$predictor) + 3*bw, bw=bw, kernel="gaussian")
density.cases <- density(rocobj$cases, from=min(rocobj$predictor) - 3 * bw,
                            to=max(rocobj$predictor) + 3*bw, bw=bw, kernel="gaussian")
smooth(rocobj, method="density", density.controls=density.controls$y, 
       density.cases=density.cases$y)
# which is roughly what is done by a simple:
smooth(rocobj, method="density")

## Not run: 
## Smoothing artificial ROC curves

rand.unif <- runif(1000, -1, 1)
rand.exp <- rexp(1000)
rand.norm <- 
rnorm(1000)

# two normals
roc.norm <- roc(controls=rnorm(1000), cases=rnorm(1000)+1, plot=TRUE)
plot(smooth(roc.norm), col="green", lwd=1, add=TRUE)
plot(smooth(roc.norm, method="density"), col="red", lwd=1, add=TRUE)
plot(smooth(roc.norm, method="fitdistr"), col="blue", lwd=1, add=TRUE)
if (requireNamespace("logcondens")) {
plot(smooth(roc.norm, method="logcondens"), col="brown", lwd=1, add=TRUE)
plot(smooth(roc.norm, method="logcondens.smooth"), col="orange", lwd=1, add=TRUE)
}
legend("bottomright", legend=c("empirical", "binormal", "density", "fitdistr",
                               "logcondens", "logcondens.smooth"), 
       col=c(par("fg"), "green", "red", "blue", "brown", "orange"), lwd=c(2, 1, 1, 1))
       
# deviation from the normality
roc.norm.exp <- roc(controls=rnorm(1000), cases=rexp(1000), plot=TRUE)
plot(smooth(roc.norm.exp), col="green", lwd=1, add=TRUE)
plot(smooth(roc.norm.exp, method="density"), col="red", lwd=1, add=TRUE)
# Wrong fitdistr: normality assumed by default
plot(smooth(roc.norm.exp, method="fitdistr"), col="blue", lwd=1, add=TRUE)
# Correct fitdistr
plot(smooth(roc.norm.exp, method="fitdistr", density.controls="normal",
            density.cases="exponential"), col="purple", lwd=1, add=TRUE)
if (requireNamespace("logcondens")) {
plot(smooth(roc.norm.exp, method="logcondens"), col="brown", lwd=1, add=TRUE)
plot(smooth(roc.norm.exp, method="logcondens.smooth"), col="orange", lwd=1, add=TRUE)
}
legend("bottomright", legend=c("empirical", "binormal", "density",
                               "wrong fitdistr", "correct fitdistr",
                               "logcondens", "logcondens.smooth"),
       col=c(par("fg"), "green", "red", "blue", "purple", "brown", "orange"), lwd=c(2, 1, 1, 1, 1))

# large deviation from the normality
roc.unif.exp <- roc(controls=runif(1000, 2, 3), cases=rexp(1000)+2, plot=TRUE)
plot(smooth(roc.unif.exp), col="green", lwd=1, add=TRUE)
plot(smooth(roc.unif.exp, method="density"), col="red", lwd=1, add=TRUE)
plot(smooth(roc.unif.exp, method="density", bw="ucv"), col="magenta", lwd=1, add=TRUE)
# Wrong fitdistr: normality assumed by default (uniform distributions not handled)
plot(smooth(roc.unif.exp, method="fitdistr"), col="blue", lwd=1, add=TRUE)
if (requireNamespace("logcondens")) {
plot(smooth(roc.unif.exp, method="logcondens"), col="brown", lwd=1, add=TRUE)
plot(smooth(roc.unif.exp, method="logcondens.smooth"), col="orange", lwd=1, add=TRUE)
}
legend("bottomright", legend=c("empirical", "binormal", "density",
                               "density ucv", "wrong fitdistr",
                               "logcondens", "logcondens.smooth"),
       col=c(par("fg"), "green", "red", "magenta", "blue", "brown", "orange"), lwd=c(2, 1, 1, 1, 1))

## End(Not run)

# 2 uniform distributions with a custom density function
unif.density <- function(x, n, from, to, bw, kernel, ...) {
  smooth.x <- seq(from=from, to=to, length.out=n)
  smooth.y <- dunif(smooth.x, min=min(x), max=max(x))
  return(smooth.y)
}
roc.unif <- roc(controls=runif(1000, -1, 1), cases=runif(1000, 0, 2), plot=TRUE)
s <- smooth(roc.unif, method="density", density=unif.density)
plot(roc.unif)
plot(s, add=TRUE, col="grey")

## Not run: 
# you can bootstrap a ROC curve smoothed with a density function:
ci(s, boot.n=100)

## End(Not run)