AutoCatalog.Featurizer Method
Definition
Important
Some information relates to prerelease product that may be substantially modified before it’s released. Microsoft makes no warranties, express or implied, with respect to the information provided here.
Overloads
| Featurizer(IDataView, ColumnInformation, String) |
Create a single featurize pipeline according to |
| Featurizer(IDataView, String, String[], String[], String[], String[], String[]) |
Create a single featurize pipeline according to |
Featurizer(IDataView, ColumnInformation, String)
Create a single featurize pipeline according to columnInformation. This function will collect all columns in columnInformation,
featurizing them using Microsoft.ML.AutoML.AutoCatalog.CatalogFeaturizer(System.String[],System.String[]), Microsoft.ML.AutoML.AutoCatalog.NumericFeaturizer(System.String[],System.String[]) or Microsoft.ML.AutoML.AutoCatalog.TextFeaturizer(System.String,System.String). And combine
them into a single feature column as output.
public Microsoft.ML.AutoML.SweepablePipeline Featurizer (Microsoft.ML.IDataView data, Microsoft.ML.AutoML.ColumnInformation columnInformation, string outputColumnName = "Features");member this.Featurizer : Microsoft.ML.IDataView * Microsoft.ML.AutoML.ColumnInformation * string -> Microsoft.ML.AutoML.SweepablePipelinePublic Function Featurizer (data As IDataView, columnInformation As ColumnInformation, Optional outputColumnName As String = "Features") As SweepablePipelineParameters
- data
- IDataView
input data.
- columnInformation
- ColumnInformation
column information.
- outputColumnName
- String
output feature column.
Returns
A SweepablePipeline for featurization.
Applies to
Featurizer(IDataView, String, String[], String[], String[], String[], String[])
Create a single featurize pipeline according to data. This function will collect all columns in data and not in excludeColumns,
featurizing them using Microsoft.ML.AutoML.AutoCatalog.CatalogFeaturizer(System.String[],System.String[]), Microsoft.ML.AutoML.AutoCatalog.NumericFeaturizer(System.String[],System.String[]) or Microsoft.ML.AutoML.AutoCatalog.TextFeaturizer(System.String,System.String). And combine
them into a single feature column as output.
public Microsoft.ML.AutoML.SweepablePipeline Featurizer (Microsoft.ML.IDataView data, string outputColumnName = "Features", string[] catelogicalColumns = default, string[] numericColumns = default, string[] textColumns = default, string[] imagePathColumns = default, string[] excludeColumns = default);member this.Featurizer : Microsoft.ML.IDataView * string * string[] * string[] * string[] * string[] * string[] -> Microsoft.ML.AutoML.SweepablePipelinePublic Function Featurizer (data As IDataView, Optional outputColumnName As String = "Features", Optional catelogicalColumns As String() = Nothing, Optional numericColumns As String() = Nothing, Optional textColumns As String() = Nothing, Optional imagePathColumns As String() = Nothing, Optional excludeColumns As String() = Nothing) As SweepablePipelineParameters
- data
- IDataView
input data.
- outputColumnName
- String
output feature column.
- catelogicalColumns
- String[]
columns that should be treated as catalog. If not specified, it will automatically infer if a column is catalog or not.
- numericColumns
- String[]
columns that should be treated as numeric. If not specified, it will automatically infer if a column is catalog or not.
- textColumns
- String[]
columns that should be treated as text. If not specified, it will automatically infer if a column is catalog or not.
- imagePathColumns
- String[]
columns that should be treated as image path. If not specified, it will automatically infer if a column is catalog or not.
- excludeColumns
- String[]
columns that won't be included when featurizing, like label