Skip to content

Commit

Permalink
Fixed typo in MK Detector Section (facebookresearch#165)
Browse files Browse the repository at this point in the history
Summary:
## Kats

Thank you for contributing! Please take a moment to review our [**contributing guidelines**](https://github.com/facebookresearch/Kats/blob/master/CONTRIBUTING.md)
to make the process easy and effective for everyone involved.

- [ ] Fork the repo and create your branch from master.
- [ ] If you've added code that should be tested, add tests.
- [ ] If you've changed APIs, update the documentation.
- [ ] Ensure the test suite passes.
- [ ] Make sure your code lints.
- [ ] If you haven't already, complete the Contributor License Agreement ("CLA").

Contributor License Agreement ("CLA")
In order to accept your pull request, we need you to submit a CLA. You only need to do this once to work on any of Facebook's open source projects.

Complete your CLA here: https://code.facebook.com/cla

`Ver: Kats v1.02`

Pull Request resolved: facebookresearch#165

Reviewed By: iamxiaodong

Differential Revision: D32006914

Pulled By: jeffhandl

fbshipit-source-id: 6990088beabe271969e04ebb38c6bc9256289ab8
  • Loading branch information
Jeffrey Handler authored and facebook-github-bot committed Nov 1, 2021
1 parent 662d12d commit 261ef24
Showing 1 changed file with 2 additions and 2 deletions.
4 changes: 2 additions & 2 deletions tutorials/kats_202_detection.ipynb
Original file line number Diff line number Diff line change
Expand Up @@ -1466,7 +1466,7 @@
"# 3. Trend detection\n",
"Trend detection tries to identify significant and prolonged changes in a time series. Rather than identifying change points, trend detection identifies windows of gradual and prolonged change.\n",
"\n",
"`MKDetector` is the trend detection algorithm that we include in Kats, which is based on the non-parametric Mann-Kendall (MK) Test. What the `MKDetector` essentially does is apply a MK to a window of fied size (specified by the `window_size` argument in the `detector` method) and return the start point of each window for which this test is statistically significant. Trend windows are detected based on the monotonicity of the increases or decreases in the time series in the window, not the magnitude of the change in the value of the time series over the window. \n",
"`MKDetector` is the trend detection algorithm that we include in Kats, which is based on the non-parametric Mann-Kendall (MK) Test. What the `MKDetector` essentially does is apply a MK to a window of fied size (specified by the `window_size` argument in the `detector` method) and return the end point of each window for which this test is statistically significant. Trend windows are detected based on the monotonicity of the increases or decreases in the time series in the window, not the magnitude of the change in the value of the time series over the window. \n",
"\n",
"The test statistic for the MK test is called the Kendall's Tau Coefficient, which ranges from -1 to 1. \n",
"* A Tau coefficient of -1 indicates a perfectly monotonic decline\n",
Expand Down Expand Up @@ -1580,7 +1580,7 @@
"metadata": {},
"source": [
"### Interpret the results\n",
"The return type of the `detector` method is `List[Tuple[TimeSeriesChangePoint, MKMetadata]]`. Each `TimeSeriesChangePoint` returned is the beginning of an increasing or decreasing trend window of duration `window_size`. In our example, we are looking for decreasing trend windows of length 20."
"The return type of the `detector` method is `List[Tuple[TimeSeriesChangePoint, MKMetadata]]`. Each `TimeSeriesChangePoint` returned is the end of an increasing or decreasing trend window of duration `window_size`. In our example, we are looking for decreasing trend windows of length 20."
]
},
{
Expand Down

0 comments on commit 261ef24

Please sign in to comment.