Thanks to visit codestin.com
Credit goes to github.com

Skip to content

Conversation

@jalopezp
Copy link

Updated the STL file to allow for missing values. There are no changes to the API but I did add a couple of _methods to allow for easier testing of the cython code. I have given a fuller description of the problem in the issue, but in summary I only modified the _est() method (in particular, no changes to ss(), ess() or how _onestp() calls these methods). I added tests that should ensure that series that do not have missing values behave exactly as before, and that series with and without missing values should be reasonably close to each other.

…in _stl.pyx to run the local regression. Updated the local regression to tolerate NAN values.
…in _stl.pyx to run the local regression. Updated the local regression to tolerate NAN values.
Updated the _est method so that it allows for
null values. The `nleft` and `nright` parameters
are no longer used if any null value is between
them. Values of Y[j] are always checked for null.
See statsmodels#8526
@jalopezp
Copy link
Author

Sorry, found a bug.

@jalopezp jalopezp closed this Feb 24, 2025
Tests now passing. The issues were with the
bounds on n, for most cases `len(X) > n`, and
`seasonal > n`.
@jalopezp jalopezp reopened this Feb 27, 2025
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants