[DOC] forecaster tutorial: multivariate forecasting, probabilistic forecasting #2041
Conversation
|
Check out this pull request on See visual diffs & provide feedback on Jupyter Notebooks. Powered by ReviewNB |
fkiraly
added this to Review in progress
in Workstream: forecasting and series transformers
via automation
fkiraly
moved this from Review in progress
to In progress
in Workstream: forecasting and series transformers
TonyBagnall
added
the
module:forecasting
|
@kejsitake could you clear the output cells in the notebook before committing changes? It'll make it easier to review the actual changes 🙂 |
fkiraly
changed the title
kejsitake
requested review from
fkiraly,
aiwalter and
TonyBagnall
as code owners
|
Thank you for the suggestion! I have cleared the cells. I see there is some difference in the notebook metadata causing the conflict(i.e. difference in python version). I am not sure if there is a preference in handling that? |
|
I think you can update to your latest version in order to resolve. |
kejsitake
force-pushed
the
tutorial
branch
5 times, most recently
from
c8dcfd5
to
b8aedd3
Compare
There was a problem hiding this comment.
Thanks for your great work, @kejsitake!
Regarding structure, I think adding the content in sections 1.2.4 and 1.2.5 is the best place, and this is the best sequence (first multivariate, then prediction intervals).
My comments revolve around being more explicit in your text. The picture that I have is being the tour guide, so you should say things like "this is the forecaster", "the input should look like this", etc. I´ve added details in my comments.
kejsitake
force-pushed
the
tutorial
branch
4 times, most recently
from
389772b
to
b515d45
Compare
* upstream/main: [ENH] Test parameter refactor: time series classifiers (sktime#2288) [ENH] Mock estimators and mock estimator generators for testing (sktime#2197) [ENH] Rename AutoARIMA from StatsForecast to StatsForecastAutoARIMA (sktime#2272) [ENH] `FeatureUnion` refactor - moved to `transformations`, tags, dunder method (sktime#2231) [DOC] - minor update to tutorials, add predict_var, predict_proba (sktime#2114) [DOC] forecaster tutorial: multivariate forecasting, probabilistic forecasting (sktime#2041) [BUG] Fix windows bug with index freq in VetorizedDF.__getitem__ (sktime#2279) refactor of TimeSeriesForestClassifier create get_test_param method. remove params from config (sktime#2277) Update .all-contributorsrc - tomiiwa (sktime#2286) [DOC] Updated pre-commit link, clearer language in Coding Style (sktime#2285) [DOC] numpydoc compliance fix of simple forecasting extension template (sktime#2284) test_param_refactor (sktime#2273) [BUG] fix conversion interval->quantiles in `BaseForecaster`, and fix `ARIMA.predict_interval` (sktime#2281)
Reference Issues/PRs
Attempt to fix #1985 and #1446
What does this implement/fix? Explain your changes.
This PR adds an explanation of the new
predict_quantilesinterface to the forecasting tutorial. It also adds a multivariate example as requested in issue #1447