search

statistikat / simPop

Simulation of Synthetic Populations for Survey Data Considering Auxiliary Information
30 stars 7 forks source link

Error in test_xgboost #19

Open alexkowa opened 2 years ago

alexkowa commented 2 years ago

Error in finalizer(env) : Inconsistent 'best_score' values between the closure state: 7590.875488 and the xgb.attr: 7590.875488

sironimo commented 2 years ago

Hi @alexkowa Thank you very much for removing the dplyr dependency and the change to tinytest. Is it just me, or is the build now faster 🚀

Regarding this error, I just looked at it, and I currently have trouble running the tests on my Windows machine. I got an error on the calibSample. If I remove the calibSample step, it runs smoothly.

Therefore a few follow up questions:

  1. What's your operating system?
  2. Which test fails? Is it the test_xgboost.R or the test_crossValidation.R?
  3. Does it run if you remove the calibSample and divide the weight rb050 by a factor (e.g., 100)?
  4. What's your XGBoost version?

Thanks and have a nice evening, Siro

sironimo commented 2 years ago

Looks like the issue is already fixed in XGBoost v1.3.0

alexkowa commented 2 years ago
  1. macOS 12.1
  2. for me only test_xgboost.R fails (test_crossValidatiion.R runs without error or warning)
  3. no, this only changes the numerical values in the error message
  4. 1.5.0.2

yes, it looks exactly like the cited issue.