Feature Tags

The Feature Analysis panel uses a comprehensive tagging system to highlight important characteristics of your data. These tags are automatically generated and use a colour-coded system to indicate different levels of significance.

The tags are organised into three main categories:

Tag Category	Tag Name	Description
Distribution	Small number of outliers	Column contains few statistical outliers
Distribution	Medium number of outliers	Column contains moderate outliers
Distribution	Significant number of outliers	Column contains many outliers
Distribution	High-skewness	Data shows significant asymmetry
Distribution	Low-variance	Data shows minimal variation
Distribution	High-variance	Data shows significant spread
Distribution	Zeros-ratio	High proportion of zero values
Pattern	Chronological-order	Values in ascending time sequence
Pattern	Reverse-chronological-order	Values in descending time sequence
Pattern	All-unique-value	Every value is unique
Pattern	Unary	Only one unique value
Pattern	High-cardinality	Large number of unique values
Balance	Multi-imbalance	Uneven distribution across categories
Balance	Multi-balance	Even distribution across categories

The Feature Tags uses a traffic light colour system to indicate different levels of significance or potential data quality concerns e.g. imbalance datasets, high/low variance features, or presence of outliers.

Color	Significance	Example Tags
Green	Good/Normal condition.	Multi-balance, Normal distribution
Yellow	Warning/Moderate concern.	Small number of outliers, Minor skewness
Red	Critical/Significant concern	Significant number of outliers, Multi-imbalance