Add frequency domain seasonal components to UnobservedComponents#3
Open
Add frequency domain seasonal components to UnobservedComponents#3
Conversation
…of the 'pseudo-inverse (np.linalg.pinv)'. Pseudo-inverse does not always preserve positive definiteness of sigma.
BUG: Changed to 'inverse of sigma (np.linalg.inv)' instead …
ENH: Improvements to factor analysis (EM algorithm)
SY: Fixed some pep8 violations in nonparametric
…kfilter REF: use full dtype in issubdtype call
We currently do not support https, and chrome shows a big warning when trying to access via https.
Ensure input is double in medcouple closes statsmodels#4243
DOC: Fixed docs as per GH4253 in stats power
DOC: MixedLM documentation issue MAINT: MixedLM pep8
MAINT: deprecation warnings, fix or silence
get rid of redundant import
small steps of algebra simplification
BUG: Fix simulation smoothing with state intercept
Easy substitutions in discrete_model
(necessary to handle all of the Py2 and Py3 / Pandas < 0.18 and Pandas >= 0.18 combinations)
ENH: Add support for RangeIndexes
…list BUG: allow list exog in get_prediction
BUG: coint use autolag in adfuller call
Previously, we had implemented the main time domain seasonal component, which
is restrictive when the periodicity is very high or the seasonality can be
more parsimoniously represented by trigonometric terms. Also, we were only
able to use a single seasonal component. This commit allows for multiple
seasonal components in the frequency domain and for mixing of the two types.
We add a test checking the state matrices for a somewhat
complicated model and update the regression test to be against
the results in the test_ucm.R file as generated by KFAS.
Finally, we added an ipython notebook as an example showcasing the new
functionality.
Also
- Change offset for regression coefficients when cycle is included.
Cycle is two states regardless of if it is stochastic. Therefore,
the offset should always be 2 instead of possibly 1 for a deterministic
cycle.
- Update the .R code for the other seasonal test to correctly initialize P1
Signed-off-by: Jordan Yoder <jordan.yoder@gmail.com>
jyoder6
force-pushed
the
ucm-complex-seas
branch
from
cdd6e23 to
24db0ff
Compare
Codecov Report
@@ Coverage Diff @@
## master #3 +/- ##
=========================================
Coverage ? 80.12%
=========================================
Files ? 560
Lines ? 84662
Branches ? 9603
=========================================
Hits ? 67832
Misses ? 14599
Partials ? 2231
Continue to review full report at Codecov.
|
This file contains hidden or 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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
15 participants
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.
trigger travis ci