Feature Type
The feature analysis process is a sophisticated type detection engine that automatically identifies the types of columns in a dataset.
Detected Type
Our custom feature detector automatically detects and assigns the following detected types:
| Type | Description |
|---|---|
| Integer | Whole number values |
| Float | Decimal/floating-point numbers |
| Currency | Decimal/floating-point numbers |
| Percentage | Values representing proportions (0-100%) |
| String | Text or categorical data |
| DateTime | Date and time values |
| Unary | Column containing only one unique value |
| Categorical | Discrete categories or labels |
| Binary | Values representing two states (e.g., True/False, 0/1) |
Note: Feature types are automatically determined and cannot be manually modified.
Feature Visualisation
The Feature Analysis panel automatically generates appropriate visualizations based on the detected feature type. Each feature is analyzed individually, with graphs tailored to best represent the data characteristics:
| Feature | Type | Description |
|---|---|---|
| Numeric (Integer/Float) | Histogram | Shows value distribution and frequency |
| Numeric (Integer/Float) | Box Plot | Displays statistical measures and outliers |
| Categorical/String | Bar Chart | Shows frequency of each unique value |
| Categorical/String | Pie Chart | Displays proportion of categories |
| DateTime | Time Series | Shows temporal patterns and trends |
| Binary | Bar Chart | Displays proportion of binary values |