feat!: add transformations and high-level forecasting API #65
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
sd2k
changed the title
sd2k
force-pushed
the
forecast-and-transformations-api
branch
from
64ea891 to
488e233
Compare
sd2k
commented
Feb 29, 2024
sd2k
force-pushed
the
forecast-and-transformations-api
branch
from
488e233 to
7b8f60d
Compare
This rather large commit adds a new `Forecaster` struct under the new `augurs-forecaster` crate, which provides a higher level API combining transforming input data, fitting and predicting. This means that models (e.g. the MSTL model) don't need to concern themselves with the potentially unlimited number of transformations that could need to happen on the input data, and instead just fit their models as normal. In doing so I needed to rework the APIs of the other models somewhat, to make them fit into a new `Fit` / `Predict` API, which makes them much easier to use (vs the old `Fit` / `Unfit` marker trait). The new APIs are similar to those used by `linfa`. The new `Predict` API also allows users to pass in a pre-allocated `Forecast` which allows for an optimization if multiple predictions are being made.
Don't use the standard Fit/Predict traits here because they make it nigh impossible to use trait objects (due to the associated types). Instead use new traits which require boxing.
sd2k
force-pushed
the
forecast-and-transformations-api
branch
from
7b8f60d to
399eb37
Compare
Closed
Closed
This was referenced Jun 5, 2024
Closed
Closed
Merged
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This rather large commit adds a new
Forecasterstruct under the newaugurs-forecastercrate, which provides a higher level APIcombining transforming input data, fitting and predicting. This means
that models (e.g. the MSTL model) don't need to concern themselves with
the potentially unlimited number of transformations that could need to
happen on the input data, and instead just fit their models as normal.
In doing so I needed to rework the APIs of the other models somewhat, to
make them fit into a new
Fit/PredictAPI, which makes them mucheasier to use (vs the old
Fit/Unfitmarker trait). The new APIsare similar to those used by
linfa.The new
PredictAPI also allows users to pass in a pre-allocatedForecastwhich allows for an optimization if multiple predictions arebeing made.