Som (SomBatchOp)

Java 类名：com.alibaba.alink.operator.batch.statistics.SomBatchOp

Python 类名：SomBatchOp

功能介绍

Self-Organized Map 算法，是一种高维数据可视化算法。

参考：https://clarkdatalabs.github.io/soms/SOM_NBA

参数说明

名称	中文名称	描述	类型	是否必须？	取值范围	默认值
vdim	向量长度	向量长度	Integer	✓
vectorCol	vector列名	vector列名	String	✓	所选列类型为 [DENSE_VECTOR, SPARSE_VECTOR, STRING, VECTOR]
xdim	x方向网格数	x方向网格数	Integer	✓
ydim	y方向网格数	y方向网格数	Integer	✓
debug	是否打开调试	是否打开调试	Boolean			false
evaluation	是否每轮评估迭代结果	是否每轮评估迭代结果	Boolean			false
learnRate	学习率	学习率	Double			0.5
numIters	迭代轮数	迭代轮数	Integer			100
sigma	neighborhood函数方差	neighborhood函数方差	Double			1.0

代码示例

Python 代码

from pyalink.alink import *

import pandas as pd

useLocalEnv(1)

df = pd.DataFrame([
    [5,2,3.5,1,'Iris-versicolor'],
    [5.1,3.7,1.5,0.4,'Iris-setosa'],
    [6.4,2.8,5.6,2.2,'Iris-virginica'],
    [6,2.9,4.5,1.5,'Iris-versicolor'],
    [4.9,3,1.4,0.2,'Iris-setosa'],
    [5.7,2.6,3.5,1,'Iris-versicolor'],
    [4.6,3.6,1,0.2,'Iris-setosa'],
    [5.9,3,4.2,1.5,'Iris-versicolor'],
    [6.3,2.8,5.1,1.5,'Iris-virginica'],
    [4.7,3.2,1.3,0.2,'Iris-setosa'],
    [5.1,3.3,1.7,0.5,'Iris-setosa'],
    [5.5,2.4,3.8,1.1,'Iris-versicolor'],
])

source = BatchOperator.fromDataframe(df, schemaStr='sepal_length double, sepal_width double, petal_length double, petal_width double, category string')

va = VectorAssemblerBatchOp().setSelectedCols(["sepal_length", "sepal_width"]) \
  .setOutputCol("features")

som = SomBatchOp()\
  .setXdim(2) \
  .setYdim(4) \
  .setVdim(2) \
  .setSigma(1.0) \
  .setNumIters(10) \
  .setVectorCol("features")

source.link(va).link(som).print()

运行结果

       meta  xidx  yidx              weights  cnt
0   4,4,2,r     0     0   6.098901,2.6216097    8
1   4,4,2,r     0     1  5.7773294,2.6960063   13
2   4,4,2,r     0     2  5.4155393,2.6422026    7
3   4,4,2,r     0     3   4.960137,2.6060686    6
4   4,4,2,r     1     0  6.3758574,2.8099446    9
5   4,4,2,r     1     1  6.1287007,2.9054923   11
6   4,4,2,r     1     2   5.422063,2.9557745    5
7   4,4,2,r     1     3  4.8237553,3.0189981   16
8   4,4,2,r     2     0   6.771961,2.9707642    8
9   4,4,2,r     2     1   6.541121,3.1317835   13
10  4,4,2,r     2     2     5.73227,3.323552    4
11  4,4,2,r     2     3  5.0270276,3.4258003   17
12  4,4,2,r     3     0   7.296618,3.1325939   11
13  4,4,2,r     3     1   6.956377,3.2311482    7
14  4,4,2,r     3     2  5.9284854,3.5446692    4
15  4,4,2,r     3     3   5.236388,3.7456865   11