Core Tools for Packages in the 'fable' Framework
Overview
Provides tools, helpers and data structures for developing models and time series functions for 'fable' and extension packages. These tools support a consistent and tidy interface for time series modelling and analysis.
New features
- Model combination now uses a joint N-way convolution that correctly accounts for the full residual covariance structure across all component models.
- Added
sum(),mean(), andweighted.mean()S3 methods formdl_tsandmdl_lstobjects, providing explicit N-way model combination with equal, uniform, or custom weights respectively. - All model arithmetic operators (
+,-,*,/) now produce a singlemodel_combinationobject with correctly implied weights. Arbitrary nested expressions such as((m1 + m2)/2 + m3)/2automatically flatten into a single joint combination.
Bug fixes
combination_ensemble()now computes inverse-variance weights using response residuals (type = "response") rather than innovation residuals. This is particularly important for models where the innovation residual variance is not comparable to the variance on the response scale, e.g. for multiplicative ETS models (#320).
Show all changes in 0.8.0Hide the rest of the changes in 0.8.0
- Replaced unintuitive behaviour of previous pairwise iterative approach for model combination (e.g.
(A + B + C)/3as((A + B) + C)/3) which used pairwise covariances that yield different results for different orders of models (#321). - Fixed computing training
accuracy()with global models.
Improvements
- Added
with_bottomargument tocoherent_smat()for inclusion of identity matrix entries for bottom level time series. Setting this to FALSE produces the A matrix (aggregated series only). - Added
"bottom"attribute forcoherent_smat()for safely identifying the index positions of bottom level series via graph coherency structures.
Breaking changes
- The
mdl_dfclass for mable data frames will be renamed tombl_df(a mable data frame). This is to avoid a naming inconsistency with themdl_tsandmdl_lst(for local and global models respectively), and free the class name for an eventualmdl_dfmulti-model class. Thembl_dfclass has been added, andmdl_dfwill be removed in the next minor release (v0.9.0).
From NEWS.md in the release
Install
Health
- OK13 OK2026-08-05
- NOTE12 OK, 1 NOTE2026-08-01
- OK14 OK2026-03-10
Documentation
- Help pages
- 88 (17 internal)includes a package overview page
- Examples that run
- 96%27 run in a plain check (19 of them under a condition) · 1 never run
- Exported functions with examples
- 23%
- Documented parameters
- 97%
- Return-value docs
- 7%
- Help pages citing references
- 8%
Downloads
Repository
- PRs merged (all time)
- 71%
- Median days to close, 50 most recently updated closed issues
- 687
- Median days to close, 50 most recently updated closed PRs
- 12
Stars over time
Forks over time
Releases over time
Issues over time
PRs over time
Commits over time
10 development-tooling and community-health practices detected across 5 families in the upstream repository
Checks run against github.com/tidyverts/fabletools on 2026-10-11.
The R CMD check workflow runs on Linux, macOS and Windows, and on R-devel
GitHub Pages site last deployed 2026-09-28 · The repository has a site deploy workflow
pkgdown site last built 2026-06-28
Show all practicesHide 1 practice
This repository shows signs of AI coding tools.
- Claude Codefirst seen 2026-10-10Pull request it wrote: opened by a project member
Dependencies
Installing fabletools adds 40 packages (21 with compiled code, 49 MB of source tarballs). Heaviest: BH 14 MB, stringi 11 MB, ggplot2 6.0 MB.
Show all 26 packagesHide all 26 packages
Code & Tests
People & History
Pick two releases to compare their code metrics.
- 0.8.0Latest
- 0.7.0
- RR 4.6.0 released · 2026-04-24
- 0.6.12026-02-16 · diff ↗
- 0.6.02026-02-09 · diff ↗
- 0.5.12025-09-01 · diff ↗
- RR 4.5.0 released · 2025-04-11
- 0.5.0
- RR 4.4.0 released · 2024-04-24
- 0.4.22024-04-22 · diff ↗
- 0.4.12024-03-02 · diff ↗
- 0.4.0
- 0.3.42023-10-12 · diff ↗
- RR 4.3.0 released · 2023-04-21
- 0.3.32023-04-04 · diff ↗ ·
removed 13 exports
parse_model, parse_model_lhs, parse_model_rhs, traverse, validate_formula, vec_cast.agg_vec, vec_cast.dcmp_ts, vec_cast.fbl_ts, vec_cast.mdl_df, vec_ptype2.agg_vec, vec_ptype2.dcmp_ts, vec_ptype2.fbl_ts, vec_ptype2.mdl_df - RR 4.2.0 released · 2022-04-22
Show 12 earlier eventsHide 12 earlier events
- 0.3.2
- RR 4.1.0 released · 2021-05-18
- 0.3.1
- 0.3.02021-01-19 · diff ↗ ·
added 8 exports
MDA, MDPV, MDV, directional_accuracy_measures, hfitted, middle_out, quantile_score, scenarios - 0.2.12020-09-04 · diff ↗ ·
removed 6 exports
dist_mv_normal, dist_normal, dist_sim, dist_unknown, new_fcdist, new_fcdist_envadded 4 exports
agg_vec, bottom_up, skill_score, top_down - 0.2.02020-06-16 · diff ↗ ·
removed 6 exports
GeomForecast, StatForecast, is_hilo, new_hilo, scale_level_continuous, scale_level_gradientadded 16 exports
RMSSE, distribution_var, is_dable, mable_vars, pinball_loss, response_vars, scaled_pinball_loss, unpack_hilo, vec_cast.agg_vec, vec_cast.dcmp_ts, vec_cast.fbl_ts, vec_cast.mdl_df, vec_ptype2.agg_vec, vec_ptype2.dcmp_ts, vec_ptype2.fbl_ts, vec_ptype2.mdl_df - RR 4.0.0 released · 2020-04-24
- 0.1.32020-03-24 · diff ↗
- 0.1.22020-01-29 · diff ↗ ·
removed 7 exports
decomposition_definition, model_definition, new_decomposition_class, new_decomposition_definition, scale_x_yearmonth, scale_x_yearquarter, scale_x_yearweekadded 1 export
MAAPE - 0.1.1
- 0.1.02019-08-08
- RR 3.6.0 released · 2019-04-26
Package metadata
- First published
- 2019-08-08
- Total releases
- 20 / 7 yrs
- License
- GPL-3 OSI
- Minimum R
- ≥ 3.1.3
- Download size
- 212 KB
Cite
Cite this package
Run in R for the authors' preferred citation:
citation("fabletools")This package has no citation file, so citation() returns this entry built from DESCRIPTION.
Cite the R Observatory
For a number measured here: a download total, a coverage figure, an archival date.
From data release v2026-10-18, which the citation names so these numbers can be found later. More on citing and the projects behind them.