缺失值填充 (Imputer)

Java 类名:com.alibaba.alink.pipeline.dataproc.Imputer

Python 类名:Imputer

功能介绍

填充缺失值,生成的模型可以用于其他数据的预处理过程

支持的填充策略包含最大值,最小值,均值和指定数值,默认为均值填充

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
selectedCols 选择的列名 计算列对应的列名列表 String[]
fillValue 填充缺失值 自定义的填充值。当strategy为value时,读取fillValue的值 String null
modelFilePath 模型的文件路径 模型的文件路径 String null
outputCols 输出结果列列名数组 输出结果列列名数组,可选,默认null String[] null
overwriteSink 是否覆写已有数据 是否覆写已有数据 Boolean false
strategy 缺失值填充规则 缺失值填充的规则,支持mean,max,min或者value。选择value时,需要读取fillValue的值 String “MEAN”, “MIN”, “MAX”, “VALUE” “MEAN”
numThreads 组件多线程线程个数 组件多线程线程个数 Integer 1
modelStreamFilePath 模型流的文件路径 模型流的文件路径 String null
modelStreamScanInterval 扫描模型路径的时间间隔 描模型路径的时间间隔,单位秒 Integer 10
modelStreamStartTime 模型流的起始时间 模型流的起始时间。默认从当前时刻开始读。使用yyyy-mm-dd hh:mm:ss.fffffffff格式,详见Timestamp.valueOf(String s) String null

代码示例

Python 代码

from pyalink.alink import *

import pandas as pd

useLocalEnv(1)

df = pd.DataFrame([
            ["a", 10.0, 100],
            ["b", -2.5, 9],
            ["c", 100.2, 1],
            ["d", -99.9, 100],
            ["a", 1.4, 1],
            ["b", -2.2, 9],
            ["c", 100.9, 1],
            [None, None, None]
])
             
colnames = ["col1", "col2", "col3"]
selectedColNames = ["col2", "col3"]

inOp = BatchOperator.fromDataframe(df, schemaStr='col1 string, col2 double, col3 double')
sinOp = StreamOperator.fromDataframe(df, schemaStr='col1 string, col2 double, col3 double')
                   
model = Imputer()\
           .setSelectedCols(selectedColNames)\
           .fit(inOp)

model.transform(inOp).print()

model.transform(sinOp).print()

StreamOperator.execute()

Java 代码

import org.apache.flink.types.Row;

import com.alibaba.alink.operator.batch.BatchOperator;
import com.alibaba.alink.pipeline.dataproc.Imputer;
import com.alibaba.alink.pipeline.dataproc.ImputerModel;
import com.alibaba.alink.operator.batch.source.MemSourceBatchOp;
import com.alibaba.alink.operator.stream.StreamOperator;
import com.alibaba.alink.operator.stream.source.MemSourceStreamOp;
import org.junit.Test;

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

public class ImputerTrainBatchOpTest {
	@Test
	public void testImputerTrainBatchOp() throws Exception {
		List <Row> df_data = Arrays.asList(
			Row.of("a", 10.0, 100),
			Row.of("b", -2.5, 9),
			Row.of("c", 100.2, 1),
			Row.of("d", -99.9, 100),
			Row.of("a", 1.4, 1),
			Row.of("b", -2.2, 9),
			Row.of("c", 100.9, 1),
			Row.of(null, null, null)
		);

		String[] selectedColNames = new String[] {"col2", "col3"};
		BatchOperator <?> inOp = new MemSourceBatchOp(df_data, "col1 string, col2 double, col3 int");
		ImputerModel imputerModel = new Imputer()
			.setSelectedCols(selectedColNames)
            .fit(inOp);
		imputerModel.transform(inOp).print();
		StreamOperator <?> sinOp = new MemSourceStreamOp(df_data, "col1 string, col2 double, col3 int");
		imputerModel.transform(inOp).print();
		StreamOperator.execute();
	}
}

运行结果

col1 col2 col3
a 10.000000 100
b -2.500000 9
c 100.200000 1
d -99.900000 100
a 1.400000 1
b -2.200000 9
c 100.900000 1
null 15.414286 31