Page History
...
The moving average calculated for Signals follows a particular calculation, and therefore will be different from the moving average calculated on a report. This is because the Signal algorithms incorporate a smoothing mechanism to limit the impact of outlier values (such as Spike, Drops) on the shape of the moving average curve. So the curves are unlikely to ever look the same, especially in the presence of one or more outliers.
Note that the Signal moving average and report moving average will be the same the first time the Signal is recorded in the analyzed time period.
...
| Anchor | ||||
|---|---|---|---|---|
|
...
The Relevance chart compares the Signals time series data (in grey, that is the metric being analyzed (represented with grey lines) with that of the baseline data (in green lines). Baseline refers to the overall dataset in which the Signal is generated, along with any access filters if applied. If Signal filtering is applied, then the baseline will include these filters as well.
Note that Relevance information is not available where the metric being analyzed contains a mix of positive and negative signed values. However if all values are either positive or negative, then relevance will be calculated.
Click on the Relevance tab to understand the relevance and importance of this Signal in relation to all your other data.
...