Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

src/metric/auc.cc:322: Check failed: auc <= local_area #10183

Open
zhuwang46 opened this issue Apr 14, 2024 · 1 comment
Open

src/metric/auc.cc:322: Check failed: auc <= local_area #10183

zhuwang46 opened this issue Apr 14, 2024 · 1 comment
Projects

Comments

@zhuwang46
Copy link

The error message occurred after changing nrounds from 32 to 33:
weight.txt

<
library("xgboost")
data(agaricus.train, package='xgboost')
data(agaricus.test, package='xgboost')
dtrain <- with(agaricus.train, xgboost::xgb.DMatrix(data, label = label))
dtest <- with(agaricus.test, xgboost::xgb.DMatrix(data, label = label))
watchlist <- list(train = dtrain, eval = dtest)
param <- list(max_depth = 2, eta = 1, nthread = 2,
objective = "binary:logitraw", eval_metric = "auc")
weight <- read.table("weight.txt", header=FALSE)[,2]
summary(weight)
xgboost::setinfo(dtrain, 'weight', weight)

nrounds=32 ok

bst <- xgboost::xgb.train(params=param, data=dtrain, nrounds = 32, watchlist=watchlist, verbose=2)

nrounds=33 crashed

bst <- xgboost::xgb.train(params=param, data=dtrain, nrounds = 33, watchlist=watchlist, verbose=2)

[1] train-auc:0.962643 eval-auc:0.960373
[2] train-auc:0.985768 eval-auc:0.979930
[3] train-auc:0.998551 eval-auc:0.998518
[4] train-auc:0.999061 eval-auc:0.998943
[5] train-auc:0.999566 eval-auc:1.000000
[6] train-auc:0.999566 eval-auc:1.000000
[7] train-auc:0.999566 eval-auc:1.000000
[8] train-auc:0.999561 eval-auc:1.000000
[9] train-auc:0.999561 eval-auc:1.000000
[10] train-auc:0.999561 eval-auc:1.000000
[11] train-auc:0.999561 eval-auc:1.000000
[12] train-auc:0.999561 eval-auc:1.000000
[13] train-auc:0.999566 eval-auc:1.000000
[14] train-auc:0.999566 eval-auc:1.000000
[15] train-auc:0.999566 eval-auc:1.000000
[16] train-auc:0.999612 eval-auc:1.000000
[17] train-auc:0.999719 eval-auc:1.000000
[18] train-auc:0.999780 eval-auc:1.000000
[19] train-auc:0.999892 eval-auc:1.000000
[20] train-auc:0.999935 eval-auc:1.000000
[21] train-auc:0.999948 eval-auc:1.000000
[22] train-auc:0.999972 eval-auc:1.000000
[23] train-auc:0.999974 eval-auc:1.000000
[24] train-auc:0.999986 eval-auc:1.000000
[25] train-auc:0.999990 eval-auc:1.000000
[26] train-auc:0.999994 eval-auc:1.000000
[27] train-auc:0.999998 eval-auc:1.000000
[28] train-auc:0.999998 eval-auc:1.000000
[29] train-auc:0.999999 eval-auc:1.000000
[30] train-auc:0.999999 eval-auc:1.000000
[31] train-auc:1.000000 eval-auc:1.000000
[32] train-auc:1.000000 eval-auc:1.000000
Error in xgb.iter.eval(bst$handle, watchlist, iteration - 1, feval) :
[12:02:28] src/metric/auc.cc:322: Check failed: auc <= local_area (1.04939e+ 07 vs. 1.04939e+07) :
Stack trace:
[bt] (0) /home/zhu/R/x86_64-pc-linux-gnu-library/4.1/xgboost/libs/xgboost. so(+0x571ec) [0x7fe35d6571ec]
[bt] (1) /home/zhu/R/x86_64-pc-linux-gnu-library/4.1/xgboost/libs/xgboost. so(+0x8beb2) [0x7fe35d68beb2]
[bt] (2) /home/zhu/R/x86_64-pc-linux-gnu-library/4.1/xgboost/libs/xgboost. so(+0x22bda5) [0x7fe35d82bda5]
[bt] (3) /home/zhu/R/x86_64-pc-linux-gnu-library/4.1/xgboost/libs/xgboost. so(XGBoosterEvalOneIter+0x204) [0x7fe35d893cb4]
[bt] (4) /home/zhu/R/x86_64-pc-linux-gnu-library/4.1/xgboost/libs/xgboost. so(XGBoosterEvalOneIter_R+0x2d4) [0x7fe35d655ec4]
[bt] (5) /usr/lib/R/lib/libR.so(+0xfa1a4) [0x7fe3652a71a4]
[bt] (6) /usr/lib/R/lib/libR.so(+0x13b1c0) [0x7fe3652e81c0]
[bt] (7) /usr/lib/R/lib/libR.so(Rf_eval+0x180) [0x7fe3652f2700]
[bt] (8) /usr/lib/R/lib/libR.so(+0x14751f) [0x7fe3652f451f]

system information:
Ubuntu 18.04
packageVersion("xgboost")
[1] ‘1.7.7.1’

@trivialfis trivialfis added this to To do in 2.1 Roadmap via automation Apr 25, 2024
@trivialfis
Copy link
Member

Apologies for the slow reply. Appears to be running into a floating error. Will try to open a PR for squashing the output auc

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
Development

No branches or pull requests

2 participants