ColumnTransformInfo
Models & Types · AiDotNet.NeuralNetworks.SyntheticData
Describes how a single original column maps into the transformed representation.
For Beginners
When the transformer processes your data, each column gets expanded into a different number of features:
- A continuous column becomes 1 (normalized value) + K (mode indicators) features
- A categorical column becomes N (one-hot categories) features
This info tracks where each original column’s features start and how wide they are.
Constructors
| Constructor | Summary |
|---|---|
ColumnTransformInfo(bool isContinuous, int index, int startOffset, int width) |
Initializes a new ColumnTransformInfo. |
Properties
| Property | Summary |
|---|---|
bool IsContinuous { get; } |
Whether this column is continuous (true) or categorical (false). |
int Index { get; } |
Index into the continuous or categorical info arrays. |
int StartOffset { get; } |
Starting offset in the transformed data vector. |
int Width { get; } |
Number of features this column occupies in the transformed representation. |