Skip to content

[ML] Wire in change detection/modelling to our univariate time series model#11

Merged
tveasey merged 13 commits intoelastic:feature/forecast-enhancements-part-2from
tveasey:feature/change-modelling-part-2
Mar 15, 2018
Merged

[ML] Wire in change detection/modelling to our univariate time series model#11
tveasey merged 13 commits intoelastic:feature/forecast-enhancements-part-2from
tveasey:feature/change-modelling-part-2

Conversation

@tveasey
Copy link
Copy Markdown
Contributor

@tveasey tveasey commented Mar 12, 2018

Description
This is work in progress, but I'm submitting to avoid this PR growing larger. This wires in change detection and starts some unit testing of CTimeSeriesModel with change points. There is some more work to be done to avoid using level shifts to try and fit other types of change points, such as scaling.

Effects
This will affect all our results, but it is work in progress so it isn't anticipated that we will test results without further changes.

Copy link
Copy Markdown

@droberts195 droberts195 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Looks good providing that it's certain that the feature branch will be ready to merge in time for 6.3 feature freeze. If it might not be ready in time then I think there's a problem that it changes the 6.3 state format.

const std::string LAST_PROPAGATION_TIME_6_3_TAG{"c"};
const std::string PERIODICITY_TEST_6_3_TAG{"d"};
const std::string CALENDAR_CYCLIC_TEST_6_3_TAG{"e"};
const std::string COMPONENTS_6_3_TAG{"f"};
Copy link
Copy Markdown

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

This PR is labelled 7.0, but doesn't this change imply that the feature branch must be merged before 6.3 feature freeze?

Copy link
Copy Markdown
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

That is a very good point! This change was mostly made before I reverted the change to bump state. I'll change this and double check that there aren't other cases where I've changed persisted state.

@tveasey tveasey changed the title Wire in change detection/modelling to our univariate time series model [ML] Wire in change detection/modelling to our univariate time series model Mar 13, 2018
Copy link
Copy Markdown

@droberts195 droberts195 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

@tveasey tveasey merged commit a468710 into elastic:feature/forecast-enhancements-part-2 Mar 15, 2018
@sophiec20 sophiec20 added the :ml label Apr 4, 2018
droberts195 pushed a commit that referenced this pull request Apr 23, 2018
… model (#11)

This wires in change detection and starts some unit testing of CTimeSeriesModel with change 
points. There is some more work to be done to avoid using level shifts to try and fit other types 
of change points, such as scaling.
@droberts195
Copy link
Copy Markdown

Removing version label as this is a feature branch PR and it causes confusion when generating release notes. (For interest this was eventually merged to 6.4 and above in #92.)

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants