等宽离散化预测 (EqualWidthDiscretizerPredictStreamOp)

参数说明

代码示例

Python 代码

from pyalink.alink import *

import pandas as pd

useLocalEnv(1)

df = pd.DataFrame([
    ["a", 1, 1.1],     
    ["b", -2, 0.9],    
    ["c", 100, -0.01], 
    ["d", -99, 100.9], 
    ["a", 1, 1.1],     
    ["b", -2, 0.9],    
    ["c", 100, -0.01], 
    ["d", -99, 100.9] 
])

batchSource =  BatchOperator.fromDataframe(df,schemaStr="f_string string, f_long long, f_double double")
streamSource =  StreamOperator.fromDataframe(df,schemaStr="f_string string, f_long long, f_double double")

trainOp = EqualWidthDiscretizerTrainBatchOp(). \
    setSelectedCols(['f_long', 'f_double']). \
    setNumBuckets(5). \
    linkFrom(batchSource)

EqualWidthDiscretizerPredictStreamOp(trainOp). \
    setSelectedCols(['f_long', 'f_double']). \
    linkFrom(streamSource). \
    print()

trainOp = EqualWidthDiscretizerTrainBatchOp().setSelectedCols(['f_long', 'f_double']). \
    setNumBucketsArray([5,3]). \
    linkFrom(batchSource)

EqualWidthDiscretizerPredictStreamOp(trainOp). \
    setSelectedCols(['f_long', 'f_double']). \
    linkFrom(streamSource). \
    print()

EqualWidthDiscretizerPredictStreamOp(trainOp). \
    setEncode("ASSEMBLED_VECTOR"). \
    setSelectedCols(['f_long', 'f_double']). \
    setOutputCols(["assVec"]). \
    linkFrom(streamSource).print()

StreamOperator.execute()

Java 代码

import org.apache.flink.types.Row;

import com.alibaba.alink.operator.batch.BatchOperator;
import com.alibaba.alink.operator.batch.feature.EqualWidthDiscretizerPredictBatchOp;
import com.alibaba.alink.operator.batch.feature.EqualWidthDiscretizerTrainBatchOp;
import com.alibaba.alink.operator.batch.source.MemSourceBatchOp;
import com.alibaba.alink.operator.stream.StreamOperator;
import com.alibaba.alink.operator.stream.feature.EqualWidthDiscretizerPredictStreamOp;
import com.alibaba.alink.operator.stream.source.MemSourceStreamOp;
import com.alibaba.alink.params.feature.HasEncodeWithoutWoe.Encode;
import org.junit.Test;

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

public class EqualWidthDiscretizerTrainBatchOpTest {
	@Test
	public void testEqualWidthDiscretizerTrainBatchOp2() throws Exception {
		List <Row> df = Arrays.asList(
			Row.of("a", 1, 1.1),
			Row.of("b", -2, 0.9),
			Row.of("c", 100, -0.01),
			Row.of("d", -99, 100.9),
			Row.of("a", 1, 1.1),
			Row.of("b", -2, 0.9),
			Row.of("c", 100, -0.01),
			Row.of("d", -99, 100.9)
		);
		BatchOperator <?> batchSource = new MemSourceBatchOp(df, "f_string string, f_long int, f_double double");

		BatchOperator <?> trainOp = new EqualWidthDiscretizerTrainBatchOp().setSelectedCols("f_long", "f_double")
			.setNumBuckets(5).linkFrom(batchSource);

		new EqualWidthDiscretizerPredictStreamOp(trainOp).setSelectedCols("f_long","f_double")
			.linkFrom(batchSource).print();

		BatchOperator trainOp2 = new EqualWidthDiscretizerTrainBatchOp().setSelectedCols("f_long", "f_double")
			.setNumBucketsArray(5,3).linkFrom(batchSource);

		new EqualWidthDiscretizerPredictStreamOp(trainOp2).setSelectedCols("f_long","f_double")
			.linkFrom(batchSource).print();

		new EqualWidthDiscretizerPredictStreamOp(trainOp2).setSelectedCols("f_long","f_double")
			.setEncode(Encode.ASSEMBLED_VECTOR)
			.setOutputCols("assVec")
			.linkFrom(batchSource).print();
	}
}

运行结果