Hi! I’m trying to cross-validate a big ordinal brms model using loo_subsample() but it crashes after a while throwing this error: Error in q[, 1L] : incorrect number of dimensions.
I could reproduce the error in one of the examples in the brms::brm documentation (I’m using brms 2.17.0 and loo 2.4.1, session info below).
fit2 <- brm(rating ~ period + carry + cs(treat),
data = inhaler, family = sratio("logit"),
prior = set_prior("normal(0,5)"), chains = 2)
loo_subsample(fit2)
loo runs fine, but it is unfeasible for my dataset due to its size.
Any ideas? Thanks!
Session info:
R version 4.0.5 (2021-03-31)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 10 x64 (build 19042)
Matrix products: default
locale:
[1] LC_COLLATE=English_United Kingdom.1252 LC_CTYPE=English_United Kingdom.1252
[3] LC_MONETARY=English_United Kingdom.1252 LC_NUMERIC=C
[5] LC_TIME=English_United Kingdom.1252
attached base packages:
[1] stats graphics grDevices datasets utils methods base
other attached packages:
[1] bayesplot_1.8.1 wesanderson_0.3.6 usethis_2.1.6 tidyr_1.1.3
[5] tidybayes_3.0.1 tibble_3.1.2 testthat_3.1.4 stringr_1.4.0
[9] stringdist_0.9.8 scales_1.1.1 rlang_1.0.4 rmarkdown_2.11.1
[13] readxl_1.3.1 purrr_0.3.4 papaja_0.1.1 tinylabels_0.2.3
[17] patchwork_1.1.1 multilex_1.0.1 mice_3.13.0 lubridate_1.7.10
[21] knitr_1.33 keyring_1.2.0 janitor_2.1.0 here_1.0.1
[25] gt_0.2.2 ggsci_2.9 ggplot2_3.3.5 forcats_0.5.1
[29] dplyr_1.0.7 conflicted_1.1.0 brms_2.17.0 Rcpp_1.0.6
[33] arrow_8.0.0 tarchetypes_0.3.0 targets_0.7.0
loaded via a namespace (and not attached):
[1] utf8_1.2.1 formr_0.9.1 tidyselect_1.1.1 lme4_1.1-27.1
[5] htmlwidgets_1.5.3 grid_4.0.5 munsell_0.5.0 codetools_0.2-18
[9] DT_0.19 miniUI_0.1.1.1 withr_2.5.0 Brobdingnag_1.2-6
[13] colorspace_2.0-2 highr_0.9 rstudioapi_0.13 stats4_4.0.5
[17] Rdpack_2.3.1 labeling_0.4.2 emmeans_1.6.3 rstan_2.21.2
[21] bit64_4.0.5 farver_2.1.0 bridgesampling_1.1-2 rprojroot_2.0.2
[25] coda_0.19-4 vctrs_0.3.8 generics_0.1.0 xfun_0.24
[29] R6_2.5.0 markdown_1.1 gamm4_0.2-6 projpred_2.0.2
[33] cachem_1.0.5 assertthat_0.2.1 promises_1.2.0.1 googlesheets4_0.3.0
[37] gtable_0.3.0 processx_3.5.2 splines_4.0.5 gargle_1.1.0
[41] broom_0.7.8 checkmate_2.0.0 inline_0.3.19 yaml_2.2.1
[45] reshape2_1.4.4 abind_1.4-5 threejs_0.3.3 crosstalk_1.1.1
[49] backports_1.2.1 httpuv_1.6.1 rsconnect_0.8.24 tensorA_0.36.2
[53] tools_4.0.5 ellipsis_0.3.2 posterior_1.1.0 RColorBrewer_1.1-2
[57] ggridges_0.5.3 plyr_1.8.6 base64enc_0.1-3 visNetwork_2.1.0
[61] ps_1.6.0 prettyunits_1.1.1 zoo_1.8-9 fs_1.5.2
[65] magrittr_2.0.1 data.table_1.14.0 ggdist_3.0.0 colourpicker_1.1.0
[69] childesr_0.2.1 googledrive_1.0.1 mvtnorm_1.1-2 matrixStats_0.60.1
[73] shinyjs_2.0.0 mime_0.11 evaluate_0.14 arrayhelpers_1.1-0
[77] xtable_1.8-4 shinystan_2.5.0 gridExtra_2.3 rstantools_2.1.1
[81] compiler_4.0.5 V8_4.2.0 crayon_1.4.1 minqa_1.2.4
[85] StanHeaders_2.21.0-7 htmltools_0.5.1.1 mgcv_1.8-33 later_1.2.0
[89] RcppParallel_5.1.4 DBI_1.1.1 MASS_7.3-53 boot_1.3-27
[93] Matrix_1.3-2 brio_1.1.3 cli_3.3.0 rbibutils_2.2.8
[97] parallel_4.0.5 igraph_1.2.6 pkgconfig_2.0.3 job_0.3.0
[101] svUnit_1.0.3 dygraphs_1.1.1.6 estimability_1.3 snakecase_0.11.0
[105] distributional_0.2.2 callr_3.7.0 digest_0.6.27 cellranger_1.1.0
[109] curl_4.3.2 shiny_1.6.0 gtools_3.9.2 commonmark_1.7
[113] nloptr_1.2.2.2 lifecycle_1.0.1 nlme_3.1-152 jsonlite_1.7.2
[117] cmdstanr_0.4.0.9000 viridisLite_0.4.0 fansi_0.5.0 pillar_1.6.1
[121] lattice_0.20-41 loo_2.4.1 fastmap_1.1.0 pkgbuild_1.3.1
[125] glue_1.6.2 xts_0.12.1 diffobj_0.3.5 shinythemes_1.2.0
[129] bit_4.0.4 stringi_1.7.8 sass_0.4.0 memoise_2.0.1
[133] renv_0.13.2