该文档涉及的组件

字符串编码训练 (StringIndexerTrainBatchOp)

Java 类名:com.alibaba.alink.operator.batch.dataproc.StringIndexerTrainBatchOp

Python 类名:StringIndexerTrainBatchOp

功能介绍

StringIndexer训练组件的作用是训练一个模型用于将单列字符串映射为整数。

如将一列映射为整数,需指定 setSelectedCol 设定。

同时,该组件支持输入多列,生成一个映射词典,通过 setSelectedCols 设定其他需要补充的列名。

特征的排列顺序支持 random,frequency_asc,frequency_desc,alphabet_asc,alphabet_desc 五种排序方法。

注意:输入多列时,所有列必须为相同格式。

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
selectedCol 选中的列名 计算列对应的列名 String 所选列类型为 [INTEGER, LONG, STRING]
modelName 模型名字 模型名字 String
selectedCols 选中的列名数组 计算列对应的列名列表 String[] 所选列类型为 [INTEGER, LONG, STRING] null
stringOrderType Token排序方法 Token排序方法 String “RANDOM”, “FREQUENCY_ASC”, “FREQUENCY_DESC”, “ALPHABET_ASC”, “ALPHABET_DESC” “RANDOM”

代码示例

Python 代码

from pyalink.alink import *

import pandas as pd

useLocalEnv(1)

df = pd.DataFrame([
    ["football", "apple"],
    ["football", "apple"],
    ["football", "apple"],
    ["basketball", "apple"],
    ["basketball", "apple"],
    ["tennis", "pair"],
    ["tennis", "pair"],
    ["pingpang", "banana"],
    ["pingpang", "banana"],
    ["baseball", "banana"]
])


data = BatchOperator.fromDataframe(df, schemaStr='f0 string,f1 string')

stringindexer = StringIndexerTrainBatchOp() \
    .setSelectedCol("f0") \
    .setSelectedCols(["f1"]) \
    .setStringOrderType("alphabet_asc")

model = stringindexer.linkFrom(data)
model.print()

Java 代码

import org.apache.flink.types.Row;

import com.alibaba.alink.operator.batch.BatchOperator;
import com.alibaba.alink.operator.batch.dataproc.StringIndexerTrainBatchOp;
import com.alibaba.alink.operator.batch.source.MemSourceBatchOp;
import org.junit.Test;

import java.util.Arrays;
import java.util.List;

public class StringIndexerTrainBatchOpTest {
	@Test
	public void testAlphabetAsc() throws Exception {
		List <Row> df = Arrays.asList(
			Row.of("football", "apple"),
			Row.of("football", "apple"),
			Row.of("football", "apple"),
			Row.of("basketball", "apple"),
			Row.of("basketball", "apple"),
			Row.of("tennis", "pair"),
			Row.of("tennis", "pair"),
			Row.of("pingpang", "banana"),
			Row.of("pingpang", "banana"),
			Row.of("baseball", "banana")
		);
		BatchOperator <?> data = new MemSourceBatchOp(df, "f0 string,f1 string");
		BatchOperator <?> stringindexer = new StringIndexerTrainBatchOp()
			.setSelectedCol("f0")
			.setSelectedCols("f1")
			.setStringOrderType("alphabet_asc");
		BatchOperator model = stringindexer.linkFrom(data);
		model.print();
	}
}

运行结果

模型表:

token token_index
pingpang 6
banana 1
baseball 2
basketball 3
pair 5
apple 0
football 4
tennis 7