General Questions

The Anomaly Score is a value between 0 and 1. The higher the number the more anomalous your data row is in the overall context of all rows submitted.

Even if you run the analysis on the same set of data, the Anomaly Score for each record, will often change slightly. This is because we use random sampling of your data and because we use some random elements to help create a faster understanding of your data. However, it is unlikely that your conclusion about whether or not a row is anomalous is going to change.


When you calculate anomaly score we store data that enables us to properly account for your usage of the service, such as the number of cells submitted, the time of day and similar. We also store some performance data to make sure you get a good service in terms of response times. We also store the name of the Excel sheet containing the data you submit. This is not the same as the filename (which we do not store), but the name of the individual worksheet. Before you change it, it’s normally called “Sheet1”. We also store the name of the range you submitted, e.g. A4:C10. When you submit data from Excel we do NOT log or store the data you submit for analysis. We keep them in memory until the calculation is over, at which time we clear them. When you submit data via our API we store the data, if your API call requests us to do this, but only then. Please see our Privacy Policy for all the details.