R reverse dependencies failures.

[trivialfis]

Dart seems to be a problem.

Package check result: OK

Changes to worse in reverse depends:

Package: nfl4th
Check: tests
New result: ERROR
    Running ‘testthat.R’ [4s/65s]
  Running the tests in ‘tests/testthat.R’ failed.
  Complete output:
    > library(testthat)
    > library(nfl4th)
    > 
    > test_check("nfl4th")
    Saving _problems/test-wrapper-64.R
    [ FAIL 1 | WARN 0 | SKIP 3 | PASS 0 ]
    
    ══ Skipped tests (3) ═══════════════════════════════════════════════════════════
    • On CRAN (3): 'test-basic.R:4:3', 'test-basic.R:15:3', 'test-wrapper.R:3:3'
    
    ══ Failed tests ════════════════════════════════════════════════════════════════
    ── Error ('test-wrapper.R:56:3'): add_2pt_probs works ──────────────────────────
    Error in `xgboost::xgb.load.raw(model)`: RAW() can only be applied to a 'raw', not a 'NULL'
    Backtrace:
         ▆
      1. ├─knitr::kable(...)
      2. ├─nfl4th::make_2pt_table_data(nfl4th::add_2pt_probs(another_play))
      3. │ ├─dplyr::select(...)
      4. │ └─tibble::tibble(...)
      5. │   └─tibble:::tibble_quos(xs, .rows, .name_repair)
      6. │     └─rlang::eval_tidy(xs[[j]], mask)
      7. └─nfl4th::add_2pt_probs(another_play)
      8.   ├─dplyr::select(...)
      9.   └─nfl4th:::get_2pt_wp(df)
     10.     ├─dplyr::ungroup(...)
     11.     ├─dplyr::mutate(...)
     12.     ├─dplyr::summarize(...)
     13.     ├─dplyr::group_by(...)
     14.     ├─dplyr::select(...)
     15.     ├─dplyr::arrange(...)
     16.     └─nfl4th:::calculate_win_probability(...)
     17.       ├─dplyr::rename(tibble::as_tibble(get_preds_wp(model_data)), vegas_home_wp = "value")
     18.       ├─tibble::as_tibble(get_preds_wp(model_data))
     19.       └─nfl4th:::get_preds_wp(model_data)
     20.         ├─stats::predict(wp_model(), as.matrix(wp_model_select(pbp)))
     21.         └─nfl4th:::wp_model()
     22.           └─nfl4th:::cached_model("wp")
     23.             └─xgboost::xgb.load.raw(model)
    
    [ FAIL 1 | WARN 0 | SKIP 3 | PASS 0 ]
    Error:
    ! Test failures.
    Execution halted

Package: tidypredict
Check: tests
New result: ERROR
    Running ‘testthat.R’ [30s/28s]
  Running the tests in ‘tests/testthat.R’ failed.
  Complete output:
    > library(testthat)
    > library(tidypredict)
    > 
    > test_check("tidypredict")
    Loading required package: Formula
    Loading required package: plotmo
    Loading required package: plotrix
    [21:33:34] WARNING: src/learner.cc:341: `booster=dart` is deprecated. Use the tree booster directly with dropout parameters like `rate_drop`, `skip_drop`, or `one_drop`.
    Saving _problems/test-model-xgboost-576.R
    Saving _problems/test-model-xgboost-603.R
    Saving _problems/test-model-xgboost-604.R
    Saving _problems/test-model-xgboost-605.R
    Saving _problems/test-model-xgboost-607.R
    [21:33:36] WARNING: src/learner.cc:824: `booster=gblinear` is deprecated and support will be removed in a future release.
    [21:33:36] WARNING: src/learner.cc:824: `booster=gblinear` is deprecated and support will be removed in a future release.
    [ FAIL 5 | WARN 1 | SKIP 170 | PASS 610 ]
    
    ══ Skipped tests (170) ═════════════════════════════════════════════════════════
    • On CRAN (89): 'test-model-catboost.R:1094:1', 'test-model-cubist.R:1:1',
      'test-model-cubist.R:41:1', 'test-model-earth.R:1:1',
      'test-model-earth.R:35:1', 'test-model-earth.R:304:1',
      'test-model-earth.R:310:1', 'test-model-earth.R:322:1',
      'test-model-glm.R:1:1', 'test-model-glm.R:39:1', 'test-model-glm.R:85:1',
      'test-model-glm.R:112:1', 'test-model-glmnet.R:1:1',
      'test-model-glmnet.R:34:1', 'test-model-glmnet.R:92:1',
      'test-model-glmnet.R:108:1', 'test-model-glmnet.R:152:1',
      'test-model-glmnet.R:163:1', 'test-model-glmnet.R:189:1',
      'test-model-glmnet.R:195:1', 'test-model-lightgbm.R:910:1',
      'test-model-lightgbm.R:1217:1', 'test-model-lightgbm.R:1229:1',
      'test-model-lightgbm.R:1242:1', 'test-model-lightgbm.R:1432:1',
      'test-model-lightgbm.R:1896:1', 'test-model-lightgbm.R:1921:1',
      'test-model-lightgbm.R:2028:1', 'test-model-lm.R:1:1',
      'test-model-lm.R:39:1', 'test-model-lm.R:83:1', 'test-model-lm.R:110:1',
      'test-model-partykit.R:1:1', 'test-model-partykit.R:28:1',
      'test-model-partykit.R:100:1', 'test-model-ranger.R:2:3',
      'test-model-ranger.R:32:3', 'test-model-ranger.R:53:3',
      'test-model-ranger.R:74:3', 'test-model-ranger.R:95:3',
      'test-model-ranger.R:116:3', 'test-model-ranger.R:136:3',
      'test-model-ranger.R:155:3', 'test-model-ranger.R:178:1',
      'test-model-ranger.R:185:3', 'test-model-ranger.R:203:3',
      'test-model-ranger.R:226:3', 'test-model-ranger.R:262:3',
      'test-model-ranger.R:287:3', 'test-model-ranger.R:307:1',
      'test-model-ranger.R:314:3', 'test-model-ranger.R:331:3',
      'test-model-ranger.R:383:1', 'test-model-ranger.R:394:3',
      'test-model-ranger.R:426:3', 'test-model-ranger.R:448:3',
      'test-model-ranger.R:470:3', 'test-model-ranger.R:494:3',
      'test-model-ranger.R:529:3', 'test-model-ranger.R:546:3',
      'test-model-ranger.R:569:3', 'test-model-rf.R:1:1', 'test-model-rf.R:34:1',
      'test-model-rf.R:79:1', 'test-model-rf.R:90:1', 'test-model-rf.R:118:1',
      'test-model-rf.R:146:1', 'test-model-rf.R:152:1', 'test-model-rf.R:237:1',
      'test-model-rf.R:243:1', 'test-model-rpart.R:31:1',
      'test-model-rpart.R:54:1', 'test-model-rpart.R:81:1',
      'test-model-rpart.R:90:1', 'test-model-rpart.R:149:1',
      'test-model-rpart.R:153:1', 'test-model-xgboost.R:753:1',
      'test-model-xgboost.R:807:1', 'test-model-xgboost.R:978:1',
      'test-model-xgboost.R:1038:1', 'test-model-xgboost.R:1147:1',
      'test-predict-interval.R:22:1', 'test-tidymodels.R:1:1',
      'test-tidypredict_test.R:24:1', 'test-tidypredict_test.R:40:1',
      'test-tidypredict_test.R:83:1', 'test-tidypredict_test.R:98:1',
      'test-tidypredict_test.R:116:1', 'test-tree.R:551:1'
    • {catboost} is not installed (81): 'test-model-catboost.R:84:3',
      'test-model-catboost.R:99:3', 'test-model-catboost.R:107:3',
      'test-model-catboost.R:129:3', 'test-model-catboost.R:141:3',
      'test-model-catboost.R:160:3', 'test-model-catboost.R:168:3',
      'test-model-catboost.R:177:3', 'test-model-catboost.R:186:3',
      'test-model-catboost.R:195:3', 'test-model-catboost.R:208:3',
      'test-model-catboost.R:241:3', 'test-model-catboost.R:275:3',
      'test-model-catboost.R:326:3', 'test-model-catboost.R:364:3',
      'test-model-catboost.R:395:3', 'test-model-catboost.R:426:3',
      'test-model-catboost.R:465:3', 'test-model-catboost.R:474:3',
      'test-model-catboost.R:483:3', 'test-model-catboost.R:496:3',
      'test-model-catboost.R:505:3', 'test-model-catboost.R:515:3',
      'test-model-catboost.R:525:3', 'test-model-catboost.R:539:3',
      'test-model-catboost.R:565:3', 'test-model-catboost.R:601:3',
      'test-model-catboost.R:633:3', 'test-model-catboost.R:665:3',
      'test-model-catboost.R:697:3', 'test-model-catboost.R:729:3',
      'test-model-catboost.R:746:3', 'test-model-catboost.R:763:3',
      'test-model-catboost.R:780:3', 'test-model-catboost.R:797:3',
      'test-model-catboost.R:833:3', 'test-model-catboost.R:869:3',
      'test-model-catboost.R:908:3', 'test-model-catboost.R:947:3',
      'test-model-catboost.R:963:3', 'test-model-catboost.R:981:3',
      'test-model-catboost.R:990:3', 'test-model-catboost.R:1014:3',
      'test-model-catboost.R:1025:3', 'test-model-catboost.R:1056:3',
      'test-model-catboost.R:1063:3', 'test-model-catboost.R:1074:3',
      'test-model-catboost.R:1104:3', 'test-model-catboost.R:1121:3',
      'test-model-catboost.R:1141:3', 'test-model-catboost.R:1161:3',
      'test-model-catboost.R:1181:3', 'test-model-catboost.R:1192:3',
      'test-model-catboost.R:1204:3', 'test-model-catboost.R:1216:3',
      'test-model-catboost.R:1233:3', 'test-model-catboost.R:1268:3',
      'test-model-catboost.R:1306:3', 'test-model-catboost.R:1317:3',
      'test-model-catboost.R:1329:3', 'test-model-catboost.R:1340:3',
      'test-model-catboost.R:1351:3', 'test-model-catboost.R:1362:3',
      'test-model-catboost.R:1401:3', 'test-model-catboost.R:1410:3',
      'test-model-catboost.R:1451:3', 'test-model-catboost.R:1493:3',
      'test-model-catboost.R:1535:3', 'test-model-catboost.R:1564:3',
      'test-model-catboost.R:1603:3', 'test-model-catboost.R:1641:3',
      'test-model-catboost.R:1685:3', 'test-model-catboost.R:1723:3',
      'test-model-catboost.R:1758:3', 'test-model-catboost.R:1803:3',
      'test-model-catboost.R:1839:3', 'test-model-catboost.R:1876:3',
      'test-model-catboost.R:1912:3', 'test-model-catboost.R:1956:3',
      'test-tidypredict_test.R:149:3', 'test-tidypredict_test.R:173:3'
    
    ══ Failed tests ════════════════════════════════════════════════════════════════
    ── Failure ('test-model-xgboost.R:576:3'): DART booster with rate_drop > 0 predictions match native predict ──
    Expected `result$alert` to be FALSE.
    Differences:
    `actual`:   TRUE 
    `expected`: FALSE
    
    ── Failure ('test-model-xgboost.R:603:3'): DART booster weight_drop is extracted correctly ──
    Expected `pm$general$booster_name` to equal "dart".
    Differences:
    `actual`:   "gbtree"
    `expected`: "dart"  
    
    ── Failure ('test-model-xgboost.R:604:3'): DART booster weight_drop is extracted correctly ──
    Expected `pm$general$weight_drop` to have type "double".
    Actual type: "NULL"
    ── Failure ('test-model-xgboost.R:605:3'): DART booster weight_drop is extracted correctly ──
    Expected `pm$general$weight_drop` to have length 4.
    Actual length: 0.
    ── Failure ('test-model-xgboost.R:607:3'): DART booster weight_drop is extracted correctly ──
    Expected `all(pm$general$weight_drop == 1)` to be FALSE.
    Differences:
    `actual`:   TRUE 
    `expected`: FALSE
    
    
    [ FAIL 5 | WARN 1 | SKIP 170 | PASS 610 ]
    Error:
    ! Test failures.
    Execution halted

[david-cortes]

Standard procedure in these situations would be to send an email to the maintainer indicated in the CRAN page of those packages:
https://cran.r-project.org/web/packages/tidypredict/index.html
https://cran.r-project.org/web/packages/nfl4th/index.html

Then give them a few weeks to make changes and reply back to CRAN or reupload.

But I see they have github repositories and you already created issues. Maybe that should be sufficient.

[david-cortes]

@trivialfis I think it's been a sufficiently long notice period by now. The email you received from CRAN should have some instructions for replying back about having notified the maintainers of affected packages.

[trivialfis]

Let me do a complete reverse deps check this time

[trivialfis]

@david-cortes I have been trying to get the reverse dependencies check to work. The latest attempt is to build a Docker image with a complete list of system dependencies and start the check from there. Still no luck, the check is still going, but the log already shows a number of "ERROR: dependencies xxx .. are not available for package yyy, please try a variant of ..`. The dependencies closure is getting much more complicated since the last time I tried to run it.

I don't have the bandwidth to triage all errors at the moment. It would be great if I could get some help on the maintenance of the CRAN package, and do let me know if you change your mind and want to take over the CRAN package maintainer position.

[trivialfis]

img.zip

[jameslamb]

The dependencies closure is getting much more complicated since the last time I tried to run it

I recall from past conversations that folks at RStudio Posit have some internal infrastructure for running reverse dependency checks. They may be willing to donate a run to {xgboost}. @simonpcouch @topepo is that still a thing?