nuget
This class implement a KMeans clustering algorithm. (请注意，实体对象的属性必须要长度一致)
KMeansAlgorithm
01 Syntax
Microsoft.VisualBasic.DataMining.KMeans.KMeansAlgorithm`1
02 Methods
| Name | Overloads | Summary |
|---|---|---|
| .ctor | 1 | |
| ClusterDataSet | 3 | Seperates a dataset into clusters or groups with similar characteristics |
| minIndex | 1 | find index for non parallel code |
03 Fields
| Name | Overloads | Summary |
|---|---|---|
| max_iters | 1 | the max iteration loop number |
04 Members
#ctor(
Boolean, Int32, Int32, Boolean, Boolean)Parameters
| Name | Type | Description |
|---|---|---|
n_threads | Boolean | 默认是使用并行化的计算代码以通过牺牲内存空间的代价来获取高性能的计算，非并行化的代码比较适合低内存的设备上面运行 |
max_iters | Int32 | the max iteration loop number |
ClusterDataSet(
IEnumerable(Of `0), Int32)Seperates a dataset into clusters or groups with similar characteristics
Remarks
if the k parameter value is greater than the element count of the source collection, then this api function will throw an exception
Parameters
| Name | Type | Description |
|---|---|---|
k | IEnumerable(Of `0) | The number of clusters or groups to form.(当这个参数值为0的时候，函数也会返回一个空集合) |
source | Int32 | An array containing data that will be clustered, the elements number must greater than 2, at least 3 elements. (里面的元素至少需要三个) |
Returns
A collection of clusters of data
ClusterDataSet(ClusterCollection(Of `0),
`0())Seperates a dataset into clusters or groups with similar characteristics
Parameters
| Name | Type | Description |
|---|---|---|
clusters | ClusterCollection(Of `0) | A collection of data clusters |
data | `0() | An array containing data to be clustered |
Returns
A collection of clusters of data
minIndex(ClusterCollection(Of `0),
`0)find index for non parallel code
Parameters
| Name | Type | Description |
|---|---|---|
clusters | ClusterCollection(Of `0) | - |
dataPoint | `0 | - |
max_iters
the max iteration loop number
ClusterDataSet(
IEnumerable(Of T), CanopySeeds, Func(Of IVector, T))