Refactor/forecast residuals fn #1223
Conversation
- past and future covariates in residuals()
- past and future covariates in residuals()
…d CONTRIBUTING.md)
…d and CONTRIBUTING.md)
…STALL.md and CONTRIBUTING.md)
| series._assert_univariate() | ||
| try: | ||
| series._assert_univariate() | ||
| except (AttributeError, TypeError): |
| @@ -0,0 +1,75 @@ | |||
| import numpy as np | |||
There was a problem hiding this comment.
+1 for the separate test file
darts/tests/utils/test_residuals.py
Outdated
| past_covariates=past_covariates, | ||
| future_covariates=future_covariates, | ||
| ) | ||
| # it seems that models can be fit with data objects that are not necessarily TimeSeries, but residuals() will |
There was a problem hiding this comment.
The only other possible type should be Sequence[TimeSeries] 😅
| target_series, | ||
| past_covariates=past_covariates, | ||
| future_covariates=future_covariates, | ||
| ) |
There was a problem hiding this comment.
Could you also add a test case where the residuals computation carries out successfully with past and/or future covariates?
|
Some of the unit tests fail, but this is because #1228 |
Codecov ReportBase: 93.71% // Head: 93.70% // Decreases project coverage by
Additional details and impacted files@@ Coverage Diff @@
## master #1223 +/- ##
==========================================
- Coverage 93.71% 93.70% -0.01%
==========================================
Files 83 83
Lines 8416 8410 -6
==========================================
- Hits 7887 7881 -6
Misses 529 529
Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here. ☔ View full report at Codecov. |
…into refactor/forecast_residuals_fn
Fixes #1127.
Summary
Other Information