{ "_id": "6a25b4d8794753ddbc3dfe31", "Package": "tabnet", "Title": "Fit 'TabNet' Models for Classification and Regression", "Version": "0.8.0.9000", "Authors@R": "c(\nperson(\"Daniel\", \"Falbel\", , \"daniel@rstudio.com\", role = \"aut\"),\nperson(, \"RStudio\", role = \"cph\"),\nperson(\"Christophe\", \"Regouby\", , \"christophe.regouby@free.fr\", role = c(\"cre\", \"ctb\")),\nperson(\"Egill\", \"Fridgeirsson\", role = \"ctb\"),\nperson(\"Philipp\", \"Haarmeyer\", role = \"ctb\"),\nperson(\"Sven\", \"Verweij\", role = \"ctb\",\ncomment = c(ORCID = \"0000-0002-5573-3952\"))\n)", "Description": "Implements the 'TabNet' model by Sercan O. Arik et al.\n(2019) with 'Coherent\nHierarchical Multi-label Classification Networks' by\nGiunchiglia et al. and\nprovides a consistent interface for fitting and creating\npredictions. It's also fully compatible with the 'tidymodels'\necosystem.", "License": "MIT + file LICENSE", "URL": "https://mlverse.github.io/tabnet/,\nhttps://github.com/mlverse/tabnet", "BugReports": "https://github.com/mlverse/tabnet/issues", "VignetteBuilder": "knitr", "Config/testthat/edition": "3", "Config/testthat/parallel": "false", "Config/testthat/start-first": "interface, explain, params", "Encoding": "UTF-8", "Roxygen": "list(markdown = TRUE)", "Language": "en-US", "Config/roxygen2/version": "8.0.0", "Config/pak/sysreqs": "cmake make libicu-dev libuv1-dev", "Repository": "https://mlverse.r-universe.dev", "Date/Publication": "2026-06-07 17:38:31 UTC", "RemoteUrl": "https://github.com/mlverse/tabnet", "RemoteRef": "HEAD", "RemoteSha": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "NeedsCompilation": "no", "Packaged": { "Date": "2026-06-07 18:09:00 UTC", "User": "root" }, "Author": "Daniel Falbel [aut],\nRStudio [cph],\nChristophe Regouby [cre, ctb],\nEgill Fridgeirsson [ctb],\nPhilipp Haarmeyer [ctb],\nSven Verweij [ctb] (ORCID: )", "Maintainer": "Christophe Regouby ", "MD5sum": "d28de188169bb31389c77ab2e157af5b", "_user": "mlverse", "_type": "src", "_file": "tabnet_0.8.0.9000.tar.gz", "_fileid": "b469ac6802ad11b08748bb358185b0a85f6366cccb41df5718f27e746ebaad12", "_filesize": 1240252, "_sha256": "b469ac6802ad11b08748bb358185b0a85f6366cccb41df5718f27e746ebaad12", "_created": "2026-06-07T18:09:00.000Z", "_published": "2026-06-07T18:13:44.331Z", "_distro": "noble", "_jobs": [ { "job": 79980537894, "time": 224, "config": "linux-devel-x86_64", "r": "4.7.0", "check": "NOTE", "artifact": "7466677498" }, { "job": 79980537893, "time": 236, "config": "linux-release-x86_64", "r": "4.6.0", "check": "NOTE", "artifact": "7466678926" }, { "job": 79980537886, "time": 143, "config": "macos-oldrel-arm64", "r": "4.5.3", "check": "NOTE", "artifact": "7466667605" }, { "job": 79980537882, "time": 177, "config": "macos-release-arm64", "r": "4.6.0", "check": "NOTE", "artifact": "7466671747" }, { "job": 79980289910, "time": 238, "config": "source", "r": "4.6.0", "check": "OK", "artifact": "7466649363" }, { "job": 79980537881, "time": 158, "config": "wasm-release", "r": "4.6.0", "check": "OK", "artifact": "7466669335" }, { "job": 79980537891, "time": 183, "config": "windows-devel", "r": "4.7.0", "check": "NOTE", "artifact": "7466672480" }, { "job": 79980537888, "time": 161, "config": "windows-oldrel", "r": "4.5.3", "check": "NOTE", "artifact": "7466670031" }, { "job": 79980537908, "time": 151, "config": "windows-release", "r": "4.6.0", "check": "NOTE", "artifact": "7466668863" } ], "_buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966", "_status": "success", "_host": "GitHub-Actions", "_upstream": "https://github.com/mlverse/tabnet", "_commit": { "id": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "author": "cregouby ", "committer": "GitHub ", "message": "Fix hierarchical ancestor missed in the transport to model (#188)\n\n* rename files for clearer organization\nadd a epoch to test\nremove processing of y in front of pretraining\n\n* augment and fix get_constr_output\nadd a proper `build_ancestor_matrix`\n\n* prune leafs in adjacency matrix extraction\n\n* add the constraint of \"No missing Values\" to hierarchical labels\ncoalesce missing labels in tests\n\n* fix hardhat_interfaces tests.\n\n* buil_ancestor_matrix is now build_ancestor_matrix_from_outcomes\nget_constr_output and nnf_multilabel_one_hot are moved in loss.R\n\n* improve `nn_mc_loss` robustness to different criterion\n\n* clone final_out before $max to prevent `variables needed for gradient computation has been modified by an inplace operation`\n\n* update news", "time": 1780853911 }, "_maintainer": { "name": "Christophe Regouby", "email": "christophe.regouby@free.fr", "login": "cregouby", "mastodon": "@cregouby@mastodon.roflcopter.fr", "description": "", "uuid": 10136115 }, "_registered": true, "_dependencies": [ { "package": "R", "version": ">= 3.6", "role": "Depends" }, { "package": "coro", "role": "Imports" }, { "package": "data.tree", "role": "Imports" }, { "package": "dials", "role": "Imports" }, { "package": "dplyr", "role": "Imports" }, { "package": "ggplot2", "role": "Imports" }, { "package": "hardhat", "version": ">= 1.3.0", "role": "Imports" }, { "package": "magrittr", "role": "Imports" }, { "package": "Matrix", "role": "Imports" }, { "package": "methods", "role": "Imports" }, { "package": "parsnip", "role": "Imports" }, { "package": "progress", "role": "Imports" }, { "package": "purrr", "role": "Imports" }, { "package": "rlang", "role": "Imports" }, { "package": "stats", "role": "Imports" }, { "package": "stringr", "role": "Imports" }, { "package": "tibble", "role": "Imports" }, { "package": "tidyr", "role": "Imports" }, { "package": "torch", "version": ">= 0.4.0", "role": "Imports" }, { "package": "tune", "role": "Imports" }, { "package": "utils", "role": "Imports" }, { "package": "vctrs", "role": "Imports" }, { "package": "withr", "role": "Imports" }, { "package": "zeallot", "role": "Imports" }, { "package": "cli", "role": "Suggests" }, { "package": "fBasics", "role": "Suggests" }, { "package": "knitr", "role": "Suggests" }, { "package": "modeldata", "role": "Suggests" }, { "package": "patchwork", "role": "Suggests" }, { "package": "quarto", "role": "Suggests" }, { "package": "recipes", "role": "Suggests" }, { "package": "rmarkdown", "role": "Suggests" }, { "package": "rsample", "role": "Suggests" }, { "package": "spelling", "role": "Suggests" }, { "package": "testthat", "version": ">= 3.0.0", "role": "Suggests" }, { "package": "tidymodels", "role": "Suggests" }, { "package": "tidyverse", "role": "Suggests" }, { "package": "vip", "role": "Suggests" }, { "package": "visdat", "role": "Suggests" }, { "package": "workflows", "role": "Suggests" }, { "package": "xgboost", "role": "Suggests" }, { "package": "yardstick", "role": "Suggests" } ], "_owner": "mlverse", "_selfowned": true, "_usedby": 0, "_updates": [ { "week": "2025-24", "n": 1 }, { "week": "2025-25", "n": 5 }, { "week": "2025-27", "n": 1 }, { "week": "2025-35", "n": 1 }, { "week": "2026-05", "n": 3 }, { "week": "2026-23", "n": 1 } ], "_tags": [ { "name": "v0.8.0", "date": "2026-01-31" } ], "_topics": [ "tabnet" ], "_stars": 116, "_contributors": [ { "user": "cregouby", "count": 220, "uuid": 10136115 }, { "user": "dfalbel", "count": 159, "uuid": 4706822 }, { "user": "christophe-regouby", "count": 86, "uuid": 167972538 }, { "user": "cmcmaster1", "count": 18, "uuid": 30722043 }, { "user": "egillax", "count": 4, "uuid": 24678081 }, { "user": "sebffischer", "count": 3, "uuid": 20975557 }, { "user": "svenvw", "count": 1, "uuid": 37927107 }, { "user": "cphaarmeyer", "count": 1, "uuid": 49557518 }, { "user": "zerweck", "count": 1, "uuid": 5678151 } ], "_userbio": { "uuid": 55406849, "type": "organization", "name": "mlverse", "description": "Open source libraries to scale Data Science" }, "_downloads": { "count": 4791, "source": "https://cranlogs.r-pkg.org/downloads/total/last-month/tabnet" }, "_devurl": "https://github.com/mlverse/tabnet", "_pkgdown": "https://mlverse.github.io/tabnet/", "_searchresults": 103, "_rbuild": "4.6.0", "_assets": [ "extra/citation.cff", "extra/citation.html", "extra/citation.json", "extra/citation.txt", "extra/contents.json", "extra/NEWS.html", "extra/NEWS.txt", "extra/readme.html", "extra/readme.md", "extra/tabnet.html", "manual.pdf" ], "_homeurl": "https://github.com/mlverse/tabnet", "_realowner": "mlverse", "_cranurl": true, "_releases": [ { "version": "0.1.0", "date": "2021-01-14" }, { "version": "0.2.0", "date": "2021-06-22" }, { "version": "0.3.0", "date": "2021-10-11" }, { "version": "0.4.0", "date": "2023-05-11" }, { "version": "0.5.0", "date": "2023-12-05" }, { "version": "0.6.0", "date": "2024-06-15" }, { "version": "0.7.0", "date": "2025-04-17" }, { "version": "0.8.0", "date": "2026-01-31" } ], "_exports": [ "%>%", "attention_width", "build_ancestor_matrix_from_outcomes", "cat_emb_dim", "check_compliant_node", "checkpoint_epochs", "decision_width", "drop_last", "encoder_activation", "entmax", "entmax15", "feature_reusage", "lr_scheduler", "mask_type", "mlp_activation", "mlp_hidden_multiplier", "momentum", "nn_aum_loss", "nn_mc_loss", "nnf_mc_loss", "nnf_multilabel_one_hot", "node_to_df", "num_independent", "num_independent_decoder", "num_shared", "num_shared_decoder", "num_steps", "optimizer", "penalty", "sparsemax", "sparsemax15", "tabnet", "tabnet_config", "tabnet_explain", "tabnet_fit", "tabnet_nn", "tabnet_pretrain", "verbose", "virtual_batch_size" ], "_help": [ { "page": "tabnet_params", "title": "Parameters for the tabnet model", "topics": [ "attention_width", "decision_width", "feature_reusage", "mask_type", "momentum", "num_independent", "num_shared", "num_steps" ] }, { "page": "autoplot.tabnet_explain", "title": "Plot tabnet_explain mask importance heatmap", "topics": [ "autoplot.tabnet_explain" ] }, { "page": "autoplot.tabnet_fit", "title": "Plot tabnet_fit model loss along epochs", "topics": [ "autoplot.tabnet_fit", "autoplot.tabnet_pretrain" ] }, { "page": "build_ancestor_matrix_from_outcomes", "title": "Build ancestor matrix aligned with observed outcome classes", "topics": [ "build_ancestor_matrix_from_outcomes" ] }, { "page": "tabnet_non_tunable", "title": "Non-tunable parameters for the tabnet model", "topics": [ "cat_emb_dim", "checkpoint_epochs", "drop_last", "encoder_activation", "lr_scheduler", "mlp_activation", "mlp_hidden_multiplier", "num_independent_decoder", "num_shared_decoder", "optimizer", "penalty", "verbose", "virtual_batch_size" ] }, { "page": "check_compliant_node", "title": "Check that Node object names are compliant", "topics": [ "check_compliant_node" ] }, { "page": "entmax15", "title": "Alpha-entmax", "topics": [ "entmax", "entmax15" ] }, { "page": "get_constr_output", "title": "Apply hierarchy constraints via max-pooling over descendants (MCM)", "topics": [ "get_constr_output" ] }, { "page": "get_tau", "title": "Optimal threshold (tau) computation for 1.5-entmax", "topics": [ "get_tau" ] }, { "page": "nn_aum_loss", "title": "AUM loss", "topics": [ "nn_aum_loss" ] }, { "page": "nn_mc_loss", "title": "Max-Constraint Margin Loss (module)", "topics": [ "nn_mc_loss" ] }, { "page": "nn_prune_head", "title": "Prune top layer(s) of a tabnet network", "topics": [ "nn_prune_head.tabnet_fit", "nn_prune_head.tabnet_pretrain" ] }, { "page": "nnf_mc_loss", "title": "Max-Constraint Margin Loss (functional)", "topics": [ "nnf_mc_loss" ] }, { "page": "nnf_multilabel_one_hot", "title": "Convert class_id tensor to binary one-hot tensor", "topics": [ "nnf_multilabel_one_hot" ] }, { "page": "node_to_df", "title": "Turn a Node object into predictor and outcome.", "topics": [ "node_to_df" ] }, { "page": "sparsemax", "title": "Sparsemax", "topics": [ "sparsemax", "sparsemax15" ] }, { "page": "tabnet", "title": "Parsnip compatible tabnet model", "topics": [ "tabnet" ] }, { "page": "tabnet_config", "title": "Configuration for TabNet models", "topics": [ "tabnet_config" ] }, { "page": "tabnet_explain", "title": "Interpretation metrics from a TabNet model", "topics": [ "tabnet_explain", "tabnet_explain.default", "tabnet_explain.model_fit", "tabnet_explain.tabnet_fit", "tabnet_explain.tabnet_pretrain" ] }, { "page": "tabnet_fit", "title": "Tabnet model", "topics": [ "tabnet_fit", "tabnet_fit.data.frame", "tabnet_fit.default", "tabnet_fit.formula", "tabnet_fit.Node", "tabnet_fit.recipe" ] }, { "page": "tabnet_nn", "title": "TabNet Model Architecture", "topics": [ "tabnet_nn" ] }, { "page": "tabnet_pretrain", "title": "Tabnet model", "topics": [ "tabnet_pretrain", "tabnet_pretrain.data.frame", "tabnet_pretrain.default", "tabnet_pretrain.formula", "tabnet_pretrain.Node", "tabnet_pretrain.recipe" ] } ], "_readme": "https://github.com/mlverse/tabnet/raw/HEAD/README.md", "_rundeps": [ "base64enc", "bit", "bit64", "bslib", "cachem", "callr", "class", "cli", "clock", "codetools", "coro", "cpp11", "crayon", "data.table", "data.tree", "desc", "diagram", "dials", "DiceDesign", "digest", "dplyr", "evaluate", "farver", "fastmap", "fontawesome", "fs", "furrr", "future", "future.apply", "GauPro", "generics", "ggplot2", "globals", "glue", "gower", "gtable", "hardhat", "highr", "hms", "htmltools", "ipred", "isoband", "jquerylib", "jsonlite", "KernSmooth", "knitr", "labeling", "lattice", "lava", "lbfgs", "lifecycle", "listenv", "lubridate", "magrittr", "MASS", "Matrix", "memoise", "mime", "mixopt", "modelenv", "nnet", "numDeriv", "otel", "parallelly", "parsnip", "pillar", "pkgconfig", "prettyunits", "processx", "prodlim", "progress", "progressr", "ps", "purrr", "R6", "rappdirs", "RColorBrewer", "Rcpp", "RcppArmadillo", "recipes", "rlang", "rmarkdown", "rpart", "rsample", "S7", "safetensors", "sass", "scales", "sfd", "shape", "slider", "sparsevctrs", "splitfngr", "SQUAREM", "stringi", "stringr", "survival", "tailor", "tibble", "tidyr", "tidyselect", "timechange", "timeDate", "tinytex", "torch", "tune", "tzdb", "utf8", "vctrs", "viridisLite", "warp", "withr", "workflows", "xfun", "yaml", "yardstick", "zeallot" ], "_vignettes": [ { "source": "tidymodels-interface.Rmd", "filename": "tidymodels-interface.html", "title": "Fitting tabnet with tidymodels", "engine": "knitr::rmarkdown", "headings": [], "created": "2020-12-18 12:02:30", "modified": "2025-04-25 15:08:54", "commits": 4 }, { "source": "Hierarchical_classification.Rmd", "filename": "Hierarchical_classification.html", "title": "Hierarchical Classification", "engine": "knitr::rmarkdown", "headings": [ "Data format", "Data preparation", "Node preparation rules for {tabnet} models", "Avoid factor predictors", "Avoid column name collision with reserved {data.tree} names", "Avoid column named level_* to avoid collision with output data.tree names", "Ensure the last hierarchy of the tree is the observation id", "Ensure there is a root level in the hierarchy", "Ensure there is no missing values in the hierarchical classes", "Model building", "Data set split", "Model diagnostic", "Model inference" ], "created": "2023-07-22 10:48:16", "modified": "2026-06-07 17:38:31", "commits": 3 }, { "source": "interpretation.Rmd", "filename": "interpretation.html", "title": "Interpretation tools", "engine": "knitr::rmarkdown", "headings": [ "Experiments", "Datasets", "Syn2", "Syn 4" ], "created": "2021-01-07 18:43:59", "modified": "2025-08-25 07:34:29", "commits": 8 }, { "source": "selfsupervised_training.Rmd", "filename": "selfsupervised_training.html", "title": "Self-supervised training and fine-tuning", "engine": "knitr::rmarkdown", "headings": [ "Data preprocessing", "Self-supervised training step", "Continuing training with supervised task", "Comparing against a model without pretraining" ], "created": "2023-07-22 10:48:16", "modified": "2025-04-15 22:24:54", "commits": 2 }, { "source": "Missing_data_predictors.Rmd", "filename": "Missing_data_predictors.html", "title": "Training a Tabnet model from missing-values dataset", "author": "Christophe Regouby", "engine": "knitr::rmarkdown", "headings": [ "Motivation", "Missing-data dataset creation", "Ames missings understanding", "While keeping some room for freedom", "Ames with missing data", "Model pretraining", "Variable importance with raw ames dataset", "Variable importance with ames_missing dataset", "Model training", "Conclusion" ], "created": "2022-02-11 17:44:42", "modified": "2025-03-30 20:53:21", "commits": 4 }, { "source": "aum_loss.Rmd", "filename": "aum_loss.html", "title": "Using ROC AUM loss for imbalanced binary classification", "engine": "knitr::rmarkdown", "headings": [ "Introduction", "How imbalance is my problem ?", "Solutions to improve imbalanced classification models", "Using the AUC metric and pr_curve() plots", "Case-weight", "ROC_AUM loss", "All together" ], "created": "2025-08-25 07:34:29", "modified": "2025-08-25 07:34:29", "commits": 1 } ], "_score": 9.916083876885487, "_indexed": true, "_nocasepkg": "tabnet", "_universes": [ "mlverse", "cregouby" ], "_binaries": [ { "r": "4.7.0", "os": "linux", "version": "0.8.0.9000", "date": "2026-06-07T18:11:53.000Z", "distro": "noble", "commit": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "fileid": "7a7ef8ff3dc87cf9538c15a593246662b5b915a4312bfe144523f8d735bf13cd", "status": "success", "check": "NOTE", "buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966" }, { "r": "4.6.0", "os": "linux", "version": "0.8.0.9000", "date": "2026-06-07T18:12:01.000Z", "distro": "noble", "commit": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "fileid": "6296bd95bd345a86c6107c95518d85fad2d5d93dbbb3aaa036797f61378ee406", "status": "success", "check": "NOTE", "buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966" }, { "r": "4.5.3", "os": "mac", "version": "0.8.0.9000", "date": "2026-06-07T18:10:54.000Z", "commit": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "fileid": "0edddb241be884b7f04a49af57febf946ca8b9e3e6e68a22ddc5896e78a15e32", "status": "success", "check": "NOTE", "buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966" }, { "r": "4.6.0", "os": "mac", "version": "0.8.0.9000", "date": "2026-06-07T18:11:16.000Z", "commit": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "fileid": "ffb29b5eb98692dc5743288d1e5d147521ecb263ff606968a9a73414e4b16c36", "status": "success", "check": "NOTE", "buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966" }, { "r": "4.6.0", "os": "wasm", "version": "0.8.0.9000", "date": "2026-06-07T18:12:00.000Z", "commit": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "fileid": "942c570292f85ae54f1810aadb0b2a72fd7b1002535dd25e5581d79ff9681800", "status": "success", "buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966" }, { "r": "4.7.0", "os": "win", "version": "0.8.0.9000", "date": "2026-06-07T18:10:47.000Z", "commit": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "fileid": "41df315a25feb6ee051d173b734a5b8c40eccef02f04fe083ce3e41da6478e60", "status": "success", "check": "NOTE", "buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966" }, { "r": "4.5.3", "os": "win", "version": "0.8.0.9000", "date": "2026-06-07T18:10:35.000Z", "commit": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "fileid": "4474182788c9c2cec52f89063dc3ba1b3ee5b5334e72d01e8065baf45afa6ef5", "status": "success", "check": "NOTE", "buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966" }, { "r": "4.6.0", "os": "win", "version": "0.8.0.9000", "date": "2026-06-07T18:10:30.000Z", "commit": "4b9b8c6cc3d6d2642eff568d24d759886466f455", "fileid": "345afde432df710a41a85dcc059f7a5fbe94f2079e65aa7f6d250fe5bbebb87e", "status": "success", "check": "NOTE", "buildurl": "https://github.com/r-universe/mlverse/actions/runs/27100523966" } ] }