nuget
The diagnostic helpers of the training DataSet object.
Diagnostics
01 Syntax
Microsoft.VisualBasic.MachineLearning.ComponentModel.StoreProcedure.Diagnostics
02 Methods
| Name | Overloads | Summary |
|---|---|---|
| CheckDataSet | 1 | Check the data integrity of a training dataset: whether the input vector size and the output vector size of each sample is consistent with the dataset definition. |
| ProjectData | 1 | Extract one property column from the sample matrix, and then evaluate the distribution of that column. |
03 Members
CheckDataSet(DataSet)
Check the data integrity of a training dataset: whether the input vector size and the output vector size of each sample is consistent with the dataset definition.
Parameters
Returns
A sequence of the LogEntry objects, each one describes a problem that was found in the data; an empty sequence means the dataset is valid.
ProjectData(
Double()(), Int32, Boolean)Extract one property column from the sample matrix, and then evaluate the distribution of that column.
Parameters
| Name | Type | Description |
|---|---|---|
matrix | Double()() | The sample matrix, in which each row is the input vector of one sample. |
index | Int32 | The column index of the target property. |
estimateQuantile | Boolean | Whether the quantile value of the distribution should be estimated? |
Returns
A SampleDistribution object which describes the distribution of the target property column.