Unexpected dates are dates that differ significantly from other observations. It may be due to normal variability or due to an error. It is similar to numeric outliers but with event variables.

How are unexpected dates detected?

The data cleaner will look for unexpected dates by assessing values which are far lower than Q1 (first quarter) and far higher than Q3 (third quarter). This is called the Tukey's fences method.

Any data greater than Q3 + 5 IQR or lower than Q1 - 5 IQR will be detected as an outlier.

How to clean data?

The process is similar to the one used for numeric outliers. Please check Numeric outliers cleaning for more information.