Loading

ZDISCORDS

gts outlier

The ZDISCORDS function detects discords in a standardized (STANDARDIZE Geo Time Series™. The discord detection algorithm is based on identifying outlier patterns in the Geo Time Series™. Outliers are patterns which differ significantly from their nearest neighbor.

The following constraints are enforced:

  • pattern length must divide window size
  • quantization scale must be a power of two from 2 to 65536
  • ZDISCORDS can only be applied to numeric Geo Time Series™ which are bucketized and filled (i.e. no bucket with missing values).

Given the ZDISCORDS expects the Geo Time Series™ it operates on to be standardized (STANDARDIZE, the discord detection algorithm is more sensitive than that of DISCORDS.

ZDISCORDS is available since version 1.0.13.

See also

Signature

Examples

NEWGTS 1 7200 // Add datapoint, with 0.1% of anomalies <% DUP 2 * PI * 360 / SIN <% RAND 0.01 < %> <% 0.001 + %> IFT NaN DUP DUP 4 ROLL ADDVALUE %> FOR [ SWAP bucketizer.last 720 1 0 ] BUCKETIZE STANDARDIZE 32 8 16 5 false 0.0 ZDISCORDS