multiROC package
Chard Liu
2019年7月28日
找ROC相关的包
Sys.setlocale('LC_ALL','C')## [1] "C"require(pkgsearch)## Loading required package: pkgsearch## Warning: package 'pkgsearch' was built under R version 3.6.1rocPkg <- pkg_search(query="ROC",size=200)
rocPkg## - "ROC" ----------------------------------- 74 packages in 0.009 seconds -
## # package version
## 1 100 pROC 1.15.3
## 2 44 caTools 1.17.1.2
## 3 18 survivalROC 1.0.3
## 4 18 PRROC 1.3.1
## 5 15 plotROC 2.2.1
## 6 14 precrec 0.10.1
## 7 14 WeightedROC 2018.10.1
## 8 13 surrosurvROC 0.1.0
## 9 13 timeROC 0.3
## 10 13 rocc 1.2
## 11 12 cvAUC 1.1.0
## 12 12 risksetROC 1.0.4
## 13 12 icsurvROC 0.1.0
## 14 12 roccv 1.2
## 15 11 plotwidgets 0.4
## 16 11 rocsvm.path 0.1.0
## 17 11 nsROC 1.1
## 18 11 ggROC 1.0
## 19 11 Comp2ROC 1.1.4
## 20 11 reportROC 3.3
## 21 10 ROCS 1.3
## 22 10 HandTill2001 0.2.12
## 23 10 ROCt 0.9.5
## 24 10 smoothROCtime 0.1.0
## 25 10 npROCRegression 1.0.5
## 26 10 sROC 0.1.2
## 27 10 TOC 0.0.4
## 28 10 HUM 1.0
## 29 10 rocNIT 1.0
## 30 10 trinROC 0.3
## 31 10 ROCwoGS 1.0
## 32 9 prognosticROC 0.7
## 33 9 rocTree 1.0.0
## 34 9 bcROCsurface 1.0.3
## 35 9 ROCR 1.0.7
## 36 8 multiROC 1.1.1
## 37 8 iMRMC 1.2.0
## 38 7 tdROC 1.0
## 39 7 nproc 2.1.4
## 40 6 ROC632 0.6
## 41 4 PresenceAbsence 1.1.9
## 42 3 hmeasure 1.0.2
## 43 3 correctedAUC 0.0.3
## 44 3 ModelGood 1.0.9
## 45 2 AUCRF 1.1
## 46 2 ezplot 0.3.1
## 47 2 ROCit 1.1.1
## 48 2 MATTOOLS 1.1
## 49 2 riskRegression 2019.1.29
## 50 2 AnaCoDa 0.1.3.0
## 51 2 AROC 1.0
## 52 2 RJafroc 1.1.0
## 53 2 logcondens 2.1.5
## 54 2 DET 2.0.1
## 55 2 Biocomb 0.4
## 56 2 WVPlots 1.1.1
## 57 2 MAMSE 0.2.1
## 58 2 classifierplots 1.3.3
## 59 2 InformationValue 1.2.3
## 60 2 EL 1.0
## 61 2 synRNASeqNet 1.0
## 62 2 mccf1 1.0
## 63 2 cubfits 0.1.3
## 64 2 liquidSVM 1.2.2.1
## 65 2 fbroc 0.4.1
## 66 1 PredictABEL 1.2.2
## 67 1 bmrm 4.1
## 68 1 intcensROC 0.1.1
## 69 1 DPpackage 1.1.7.4
## 70 1 bimixt 1.0
## 71 1 nlcv 0.3.5
## 72 1 RcmdrPlugin.EZR 1.40
## 73 1 r4lineups 0.1.1
## 74 1 psfmi 0.1.0
## by
## Xavier Robin
## ORPHANED
## Paramita Saha-Chaudhuri<U+000a><paramita.sahachaudhuri.work@gmail.com>
## Jan Grau
## Michael C. Sachs
## Takaya Saito
## Toby Dylan Hocking
## Yunro Chung
## Paul Blanche
## Martin Lauss
## Erin LeDell
## Paramita Saha-Chaudhuri<U+000a><paramita.sahachaudhuri.work@gmail.com>
## Yunro Chung
## Ben Sherwood
## January Weiner
## Seung Jun Shin
## Sonia Perez Fernandez
## Honglong Wu
## Ana C. Braga
## Zhicheng Du<dgdzc@hotmail.com>
## Tianwei Yu<tyu8@emory.edu>
## Andreas Dominik Cullmann
## Y. Foucher
## Susana Diaz-Coto
## Maria Xose Rodriguez-Alvarez
## Xiao-Feng Wang
## Al<U+00ED> Santacruz
## Natalia Novoselova
## Zhicheng Du<dgdzc@hotmail.com>
## Reinhard Furrer
## Chong Wang
## Y. Foucher
## Sy Han Chiou
## Khanh To Duc
## Tobias Sing
## Runmin Wei
## Brandon Gallas
## Cai Wu
## Yang Feng
## Y. Foucher
## Elizabeth Freeman
## Christoforos Anagnostopoulos
## Weiliang Qiu
## Thomas A. Gerds
## Victor Urrea
## Wojtek Kostelecki
## Md Riaz Ahmed Khan
## steven mosher
## Thomas Alexander Gerds
## Cedric Landerer
## Maria Xose Rodriguez-Alvarez
## Dev Chakraborty
## Kaspar Rufibach
## "Garc<U+00ED>a-R<U+00F3>denas, <U+00C1>lvaro"
## Natalia Novoselova
## John Mount
## Jean-Francois Plante
## Huw Campbell
## Selva Prabhakaran
## Edmunds Cers
## Luciano Garofano
## Chang Cao
## Wei-Chen Chen
## ORPHANED
## Erik Peter
## Suman Kundu
## Julien Prados
## Jiaxing Lin<jiaxing.lin@duke.edu>
## ORPHANED
## Michelle Winerip
## Laure Cougnaud
## Yoshinobu Kanda
## Colin Tredoux
## Martijn Heymans
## @
## 7d
## 5M
## 7y
## 1y
## 1y
## 4M
## 10M
## 11M
## 4y
## 9y
## 5y
## 7y
## 8M
## 3M
## 3y
## 10M
## 1y
## 6y
## 3y
## 4M
## 3y
## 3y
## 2y
## 9M
## 2y
## 7y
## 4y
## 6y
## 3y
## 2M
## 7y
## 6y
## 4M
## 10M
## 4y
## 1y
## 3M
## 3y
## 8M
## 6y
## 7y
## 5M
## 3y
## 5y
## 7y
## 8d
## 6M
## 7y
## 6M
## 3M
## 1y
## 9M
## 3y
## 1M
## 1y
## 4d
## 3y
## 2y
## 3y
## 8y
## 4y
## 1y
## 2y
## 5M
## 4M
## 5y
## 4M
## 1y
## 2y
## 4y
## 1y
## 3M
## 1y
## 2Mhead(rocPkg)## - "ROC" ----------------------------------- 74 packages in 0.009 seconds -
## # package version
## 1 100 pROC 1.15.3
## 2 44 caTools 1.17.1.2
## 3 18 survivalROC 1.0.3
## 4 18 PRROC 1.3.1
## 5 15 plotROC 2.2.1
## 6 14 precrec 0.10.1
## by
## Xavier Robin
## ORPHANED
## Paramita Saha-Chaudhuri<U+000a><paramita.sahachaudhuri.work@gmail.com>
## Jan Grau
## Michael C. Sachs
## Takaya Saito
## @
## 7d
## 5M
## 7y
## 1y
## 1y
## 4MmultiROC包
require(multiROC)## Loading required package: multiROC## Warning: package 'multiROC' was built under R version 3.6.1data(iris)
head(iris)## Sepal.Length Sepal.Width Petal.Length Petal.Width Species
## 1 5.1 3.5 1.4 0.2 setosa
## 2 4.9 3.0 1.4 0.2 setosa
## 3 4.7 3.2 1.3 0.2 setosa
## 4 4.6 3.1 1.5 0.2 setosa
## 5 5.0 3.6 1.4 0.2 setosa
## 6 5.4 3.9 1.7 0.4 setosaset.seed(123456)
total_number <- nrow(iris)
train_idx <- sample(total_number, round(total_number*0.6))
## 随机抽样分train,test组
train_df <- iris[train_idx, ]
test_df <- iris[-train_idx, ]Random forest法
rf_res <- randomForest::randomForest(Species~., data = train_df, ntree = 100)
rf_res##
## Call:
## randomForest(formula = Species ~ ., data = train_df, ntree = 100)
## Type of random forest: classification
## Number of trees: 100
## No. of variables tried at each split: 2
##
## OOB estimate of error rate: 7.78%
## Confusion matrix:
## setosa versicolor virginica class.error
## setosa 30 0 0 0.0000000
## versicolor 0 29 3 0.0937500
## virginica 0 4 24 0.1428571rf_pred <- predict(rf_res, test_df, type = 'prob')
rf_pred <- data.frame(rf_pred)
colnames(rf_pred) <- paste(colnames(rf_pred), "_pred_RF")logistic回归模型
mn_res <- nnet::multinom(Species ~., data = train_df)## # weights: 18 (10 variable)
## initial value 98.875106
## iter 10 value 12.524348
## iter 20 value 5.495452
## iter 30 value 5.352345
## iter 40 value 5.304605
## iter 50 value 5.251277
## iter 60 value 5.250840
## final value 5.250449
## convergedmn_pred <- predict(mn_res, test_df, type = 'prob')
mn_pred <- data.frame(mn_pred)
colnames(mn_pred) <- paste(colnames(mn_pred), "_pred_MN")整合预测值与真实值
## 设置真实值1为TRUE
true_label <- dummies::dummy(test_df$Species, sep = ".")## Warning in model.matrix.default(~x - 1, model.frame(~x - 1), contrasts =
## FALSE): non-list contrasts argument ignoredtrue_label <- data.frame(true_label)
colnames(true_label) <- gsub(".*?\\.", "", colnames(true_label))
colnames(true_label) <- paste(colnames(true_label), "_true")
## 整合
final_df <- cbind(true_label, rf_pred, mn_pred)
head(final_df)## setosa _true versicolor _true virginica _true setosa _pred_RF
## 6 1 0 0 1.00
## 7 1 0 0 1.00
## 17 1 0 0 1.00
## 18 1 0 0 1.00
## 19 1 0 0 0.98
## 22 1 0 0 1.00
## versicolor _pred_RF virginica _pred_RF setosa _pred_MN
## 6 0.00 0 1
## 7 0.00 0 1
## 17 0.00 0 1
## 18 0.00 0 1
## 19 0.02 0 1
## 22 0.00 0 1
## versicolor _pred_MN virginica _pred_MN
## 6 6.639968e-11 4.388237e-33
## 7 5.584571e-09 9.246104e-31
## 17 1.097355e-13 3.009961e-37
## 18 1.539783e-10 4.700440e-33
## 19 4.592017e-11 6.295323e-34
## 22 8.766548e-11 8.012250e-33multiROC
force_diag=T, 则TPR，FPR的值转换为0-1之间
roc_res <- multi_roc(final_df, force_diag=T)## Warning in regularize.values(x, y, ties, missing(ties)): collapsing to
## unique 'x' values
## Warning in regularize.values(x, y, ties, missing(ties)): collapsing to
## unique 'x' values
## Warning in regularize.values(x, y, ties, missing(ties)): collapsing to
## unique 'x' values
## Warning in regularize.values(x, y, ties, missing(ties)): collapsing to
## unique 'x' values
## Warning in regularize.values(x, y, ties, missing(ties)): collapsing to
## unique 'x' values
## Warning in regularize.values(x, y, ties, missing(ties)): collapsing to
## unique 'x' valuesplot
多出的MacroROC,MicroROC为计算的新的AUC值 data为plot_roc_df, x为1-特异度, y为敏感度
## 将数据转换为ggplot喜欢的格式
plot_roc_df <- plot_roc_data(roc_res)
head(plot_roc_df)## Specificity Sensitivity Group AUC Method
## 1 1 0.00 setosa 1 RF
## 2 1 0.05 setosa 1 RF
## 3 1 0.10 setosa 1 RF
## 4 1 0.15 setosa 1 RF
## 5 1 0.20 setosa 1 RF
## 6 1 0.25 setosa 1 RF## 绘图
require(ggplot2)## Loading required package: ggplot2ggplot(plot_roc_df, aes(x = 1-Specificity, y=Sensitivity)) +
geom_path(aes(color = Group, linetype=Method), size=1.5) +
geom_segment(aes(x = 0, y = 0, xend = 1, yend = 1),
colour='grey', linetype = 'dotdash') +
theme_bw() +
theme(plot.title = element_text(hjust = 0.5),
legend.justification=c(1, 0), legend.position=c(.95, .05),
legend.title=element_blank(),
legend.background = element_rect(fill=NULL, size=0.5,
linetype="solid", colour ="black"))
总结
- multiROC包适用于比较不同的模型应用于诊断的ROC比较
- multiROC可计算多分类诊断模型