Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

[ENH] An ensemble type forecaster for panel/hierarchical data #2764

Closed
ciaran-g opened this issue Jun 8, 2022 · 3 comments · Fixed by #3905
Closed

[ENH] An ensemble type forecaster for panel/hierarchical data #2764

ciaran-g opened this issue Jun 8, 2022 · 3 comments · Fixed by #3905
Labels
enhancement Adding new functionality feature request New feature or request good first issue Good for newcomers module:forecasting forecasting module: forecasting, incl probabilistic and hierarchical forecasting

Comments

@ciaran-g
Copy link
Contributor

ciaran-g commented Jun 8, 2022

Is your feature request related to a problem? Please describe.
For forecasting in a hierarchy it would be nice to be able to use different forecasters for each individual node

Describe the solution you'd like
A class similar to ColumnEnsembleForecaster but instead of applying a different univariate forecaster per-column, it applies a different univariate forecaster per-level on panel-type data

Describe alternatives you've considered
One option is to run a loop or something but it would be difficult to adapt this into #2157
@ciaran-g ciaran-g added the enhancement Adding new functionality label Jun 8, 2022
@ciaran-g ciaran-g changed the title [ENH] - An ensemble type forecaster for panel/hierarchical data [ENH] An ensemble type forecaster for panel/hierarchical data Jun 8, 2022
@fkiraly
Copy link
Collaborator

fkiraly commented Jun 15, 2022

We should also check whether VectorizedDF can be used for this productively. Probably yes?

@fkiraly fkiraly added feature request New feature or request good first issue Good for newcomers module:forecasting forecasting module: forecasting, incl probabilistic and hierarchical forecasting labels Jun 15, 2022
@VyomkeshVyas
Copy link
Contributor

Is your feature request related to a problem? Please describe. For forecasting in a hierarchy it would be nice to be able to use different forecasters for each individual node

Describe the solution you'd like A class similar to ColumnEnsembleForecaster but instead of applying a different univariate forecaster per-column, it applies a different univariate forecaster per-level on panel-type data

Describe alternatives you've considered One option is to run a loop or something but it would be difficult to adapt this into #2157

Hi @ciaran-g , @fkiraly I am currently working on this problem.

@fkiraly
Copy link
Collaborator

fkiraly commented Nov 7, 2022

excellent, @VyomkeshVyas! Looking forward to your PR.

@VyomkeshVyas VyomkeshVyas mentioned this issue Dec 8, 2022
Merged
11 tasks
fkiraly pushed a commit that referenced this issue Jan 28, 2023
@VyomkeshVyas
[ENH] HierarchyEnsembleForecaster for level- or node-wise applicati…
eb2ca69 
…on of forecasters on panel/hierarchical data (#3905)

Fixes #2764 , adds a new compositor, `HierarchyEnsembleForecaster`

`HierarchyEnsembleForecaster` aggregates panel-type data and applies different univariate forecasters on the aggregated data by each hierarchical level/node. For aggregation, it employs sktime's bulit-in 'Aggregator' class.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
enhancement Adding new functionality feature request New feature or request good first issue Good for newcomers module:forecasting forecasting module: forecasting, incl probabilistic and hierarchical forecasting
Projects
None yet
Milestone
No milestone
3 participants
@fkiraly @ciaran-g @VyomkeshVyas