You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: docs/en/06-advanced/06-tdgpt/06-anomaly-detection/index.md
+1-1Lines changed: 1 addition & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -30,7 +30,7 @@ algo=expr1
30
30
"}
31
31
```
32
32
33
-
1.`column_name`: The data column in which to detect anomalies. Specify only one column per query. The data type of the column must be numerical; string types such as NCHAR are not supported. Functions are not supported.
33
+
1.`column_name(s)`: The data column(s) in which to detect anomalies. The data type of the column(s) must be numerical; string types such as NCHAR are not supported. Functions are not supported. Starting from version 3.4.1.0, multi-column data input is supported for anomaly detection. For models that only handle a single column, only the first column will be used, and any additional columns will be ignored automatically.
34
34
2.`options`: The parameters for anomaly detection. Enter parameters in key=value format, separating multiple parameters with a comma (,). It is not necessary to use quotation marks or escape characters. Only ASCII characters are supported. For example: `algo=ksigma,k=2` indicates that the anomaly detection algorithm is k-sigma and the k value is 2.
35
35
3. You can use the results of anomaly detection as the inner part of a nested query. The same functions are supported as in other windowed queries.
36
36
4. White noise checking is performed on the input data by default. If the input data is white noise, no results are returned.
`execute` is the core method of anomaly detection algorithms. Before calling this method, the framework configures the historical time-series data used for anomaly detection in the `self.list` object attribute.
8
+
`execute` is the core method of anomaly detection algorithms. Before calling this method, the framework configures the historical time-series data used for anomaly detection in the `self.list` member variable.
9
+
10
+
Starting from version 3.4.1.0, a new Python member variable, `self.input_data_lists`, has been added to all anomaly detection classes inherited from `AbstractAnomalyDetectionService`. This object is an array of arrays (nested array) used to store all columns of multi-column input data for anomaly detection. For backward compatibility, the framework still keeps `self.list` as the first input column, while `self.input_data_lists` contains every input column. In `execute`, use `self.list` for single-column logic or backward-compatible implementations, and use `self.input_data_lists` when your algorithm needs access to multiple columns.
| syncAssignedCheckAppliedGap| After 3.4.1.0 | Supported, effective immediately | Threshold for the gap between peer's applied index and commit index before an assigned leader steps down in dual-replica mode. The assigned leader only steps down when the gap is within this value, preventing the peer from entering restoring state. 0 means no gap check (immediate step down). Range: 0-10000, default: 20. Internal parameter, for debugging synchronization module |
0 commit comments