Anomaly Detection - How can I show what attributes contribute to a dip/spike on a line graph?

[bvy]

Recently Power BI introduced Anomaly Detection on its native line graph. I’m interested in what other visuals offer similar functionality.  

For example, suppose we’re tracking sales over time, and we see a spike. The report user can try slicing and dicing through different products, customers, promotions, times of day, etc. to determine what that spike was attributed to. But the line graph with anomaly detection plants a marker there and when you click on it, it makes suggestions -- e.g. with XX% confidence we think the spike was attributed to this product.

This is great, but it puts this information in the Anomalies pane which users find a little cumbersome and not the most intuitive. It would be nice if that same information could be dumped to a table, for example. Are there workarounds to this?

Basically it would be nice to know, for outliers on a line graph, what dimensions are responsible for the anomaly.

Thanks.

[Anonymous · Accepted solution]

Hi @bvy ,

Anomaly detection helps you enhance your line charts by automatically detecting anomalies in your time series data. It also provides explanations for the anomalies to help with root cause analysis. 

Besides detecting anomalies, you can also automatically explain the anomalies the data. When you select the anomaly, Power BI runs an analysis across fields in your data model to figure out possible explanations. It gives you a natural language explanation of the anomaly, and factors associated with that anomaly, sorted by its explanatory strength. 

For more details, please see:

Anomaly detection 

Anomaly detection tutorial - Power BI | Microsoft Docs

Hope it helps,

Community Support Team _ Caitlyn

If this post helps then please consider Accept it as the solution to help the other members find it more quickly.