USB連接器-USB Micro/Mini Connector

主要從事各式USB連接器、插座、開關、線材之研發製造及銷售，以高品質、合理的價位、快速的交貨期及高滿意度的服務行銷國際。USB 2.0版 & USB3.0版 & USB3.1版連接器、USB AM連接器、USB AF連接器、USB BF連接器、USB BM連接器、MICRO USB連接器、MINI USB連接器、IEEE 1394連接器、HDMI連接器等各式電子零件產品。

Alink漫談(八) : 二分類評估 AUC、K-S、PRC、Precision、Recall、LiftChart 如何實現

由 · 已發表 · 已更新

Alink漫談(八) : 二分類評估 AUC、K-S、PRC、Precision、Recall、LiftChart 如何實現

目錄

  • Alink漫談(八) : 二分類評估 AUC、K-S、PRC、Precision、Recall、LiftChart 如何實現
    • 0x00 摘要
    • 0x01 相關概念
    • 0x02 示例代碼
      • 2.1 主要思路
    • 0x03 批處理
      • 3.1 EvalBinaryClassBatchOp
      • 3.2 BaseEvalClassBatchOp
        • 3.2.0 調用關係綜述
        • 3.2.1 calLabelPredDetailLocal
          • 3.2.1.1 flatMap
          • 3.2.1.2 reduceGroup
          • 3.2.1.3 mapPartition
        • 3.2.2 ReduceBaseMetrics
        • 3.2.3 SaveDataAsParams
        • 3.2.4 計算混淆矩陣
          • 3.2.4.1 原始矩陣
          • 3.2.4.2 計算標籤
          • 3.2.4.3 具體代碼
    • 0x04 流處理
      • 4.1 示例
        • 4.1.1 主類
        • 4.1.2 TimeMemSourceStreamOp
        • 4.1.3 Source
      • 4.2 BaseEvalClassStreamOp
        • 4.2.1 PredDetailLabel
        • 4.2.2 AllDataMerge
        • 4.2.3 SaveDataStream
        • 4.2.4 Union
          • 4.2.4.1 allOutput
        • 4.2.4.2 windowOutput
    • 0xFF 參考

0x00 摘要

Alink 是阿里巴巴基於實時計算引擎 Flink 研發的新一代機器學習算法平台，是業界首個同時支持批式算法、流式算法的機器學習平台。二分類評估是對二分類算法的預測結果進行效果評估。本文將剖析Alink中對應代碼實現。

0x01 相關概念

如果對本文某些概念有疑惑，可以參見之前文章 [白話解析] 通過實例來梳理概念 ：準確率 (Accuracy)、精準率(Precision)、召回率(Recall) 和 F值(F-Measure)

0x02 示例代碼

public class EvalBinaryClassExample {

    AlgoOperator getData(boolean isBatch) {
        Row[] rows = new Row[]{
                Row.of("prefix1", "{\"prefix1\": 0.9, \"prefix0\": 0.1}"),
                Row.of("prefix1", "{\"prefix1\": 0.8, \"prefix0\": 0.2}"),
                Row.of("prefix1", "{\"prefix1\": 0.7, \"prefix0\": 0.3}"),
                Row.of("prefix0", "{\"prefix1\": 0.75, \"prefix0\": 0.25}"),
                Row.of("prefix0", "{\"prefix1\": 0.6, \"prefix0\": 0.4}")
        };

        String[] schema = new String[]{"label", "detailInput"};

        if (isBatch) {
            return new MemSourceBatchOp(rows, schema);
        } else {
            return new MemSourceStreamOp(rows, schema);
        }
    }

    public static void main(String[] args) throws Exception {
        EvalBinaryClassExample test = new EvalBinaryClassExample();
        BatchOperator batchData = (BatchOperator) test.getData(true);

        BinaryClassMetrics metrics = new EvalBinaryClassBatchOp()
                .setLabelCol("label")
                .setPredictionDetailCol("detailInput")
                .linkFrom(batchData)
                .collectMetrics();

        System.out.println("RocCurve:" + metrics.getRocCurve());
        System.out.println("AUC:" + metrics.getAuc());
        System.out.println("KS:" + metrics.getKs());
        System.out.println("PRC:" + metrics.getPrc());
        System.out.println("Accuracy:" + metrics.getAccuracy());
        System.out.println("Macro Precision:" + metrics.getMacroPrecision());
        System.out.println("Micro Recall:" + metrics.getMicroRecall());
        System.out.println("Weighted Sensitivity:" + metrics.getWeightedSensitivity());
    }
}

程序輸出

RocCurve:([0.0, 0.0, 0.0, 0.5, 0.5, 1.0, 1.0],[0.0, 0.3333333333333333, 0.6666666666666666, 0.6666666666666666, 1.0, 1.0, 1.0])
AUC:0.8333333333333333
KS:0.6666666666666666
PRC:0.9027777777777777
Accuracy:0.6
Macro Precision:0.3
Micro Recall:0.6
Weighted Sensitivity:0.6

在 Alink 中，二分類評估有批處理，流處理兩種實現，下面一一為大家介紹（ Alink 複雜之一在於大量精細的數據結構，所以下文會大量打印程序中變量以便大家理解）。

2.1 主要思路

  • 把 [0,1] 分成假設 100000個桶(bin)。所以得到positiveBin / negativeBin 兩個100000的數組。

  • 根據輸入給positiveBin / negativeBin賦值。positiveBin就是 TP + FP，negativeBin就是 TN + FN。這些是後續計算的基礎。

  • 遍歷bins中每一個有意義的點，計算出totalTrue和totalFalse，並且在每一個點上計算該點的混淆矩陣，tpr，以及rocCurve，recallPrecisionCurve，liftChart在該點對應的數據；

  • 依據曲線內容計算並且存儲 AUC/PRC/KS

具體後續還有詳細調用關係綜述。

0x03 批處理

3.1 EvalBinaryClassBatchOp

EvalBinaryClassBatchOp是二分類評估的實現，功能是計算二分類的評估指標(evaluation metrics)。

輸入有兩種：

  • label column and predResult column
  • label column and predDetail column。如果有predDetail，則predResult被忽略

我們例子中 "prefix1" 就是 label，"{\"prefix1\": 0.9, \"prefix0\": 0.1}" 就是 predDetail

Row.of("prefix1", "{\"prefix1\": 0.9, \"prefix0\": 0.1}")

具體類摘錄如下：

public class EvalBinaryClassBatchOp extends BaseEvalClassBatchOp<EvalBinaryClassBatchOp> implements BinaryEvaluationParams <EvalBinaryClassBatchOp>, EvaluationMetricsCollector<BinaryClassMetrics> {
  
	@Override
	public BinaryClassMetrics collectMetrics() {
		return new BinaryClassMetrics(this.collect().get(0));
	}  
}

可以看到，其主要工作都是在基類BaseEvalClassBatchOp中完成，所以我們會首先看BaseEvalClassBatchOp。

3.2 BaseEvalClassBatchOp

我們還是從 linkFrom 函數入手，其主要是做了幾件事：

  • 獲取配置信息
  • 從輸入中提取某些列：”label”，”detailInput”
  • calLabelPredDetailLocal會按照partition分別計算evaluation metrics
  • 綜合reduce上述計算結果
  • SaveDataAsParams函數會把最終數值輸入到 output table

具體代碼如下

@Override
public T linkFrom(BatchOperator<?>... inputs) {
    BatchOperator<?> in = checkAndGetFirst(inputs);
    String labelColName = this.get(MultiEvaluationParams.LABEL_COL);
    String positiveValue = this.get(BinaryEvaluationParams.POS_LABEL_VAL_STR);

    // Judge the evaluation type from params.
    ClassificationEvaluationUtil.Type type = ClassificationEvaluationUtil.judgeEvaluationType(this.getParams());

    DataSet<BaseMetricsSummary> res;
    switch (type) {
        case PRED_DETAIL: {
            String predDetailColName = this.get(MultiEvaluationParams.PREDICTION_DETAIL_COL);
            // 從輸入中提取某些列："label"，"detailInput" 
            DataSet<Row> data = in.select(new String[] {labelColName, predDetailColName}).getDataSet();
            // 按照partition分別計算evaluation metrics
            res = calLabelPredDetailLocal(data, positiveValue, binary);
            break;
        }
        ......
    }

    // 綜合reduce上述計算結果
    DataSet<BaseMetricsSummary> metrics = res
        .reduce(new EvaluationUtil.ReduceBaseMetrics());

    // 把最終數值輸入到 output table
    this.setOutput(metrics.flatMap(new EvaluationUtil.SaveDataAsParams()),
        new String[] {DATA_OUTPUT}, new TypeInformation[] {Types.STRING});

    return (T)this;
}

// 執行中一些變量如下
labelColName = "label"
predDetailColName = "detailInput"  
type = {ClassificationEvaluationUtil$Type@2532} "PRED_DETAIL"
binary = true
positiveValue = null

3.2.0 調用關係綜述

因為後續代碼調用關係複雜，所以先給出一個調用關係：

  • 從輸入中提取某些列：”label”，”detailInput”，in.select(new String[] {labelColName, predDetailColName}).getDataSet()。因為可能輸入還有其他列，而只有某些列是我們計算需要的，所以只提取這些列。
  • 按照partition分別計算evaluation metrics，即調用 calLabelPredDetailLocal(data, positiveValue, binary);
    • flatMap會從label列和prediction列中，取出所有labels（注意是取出labels的名字 ），發送給下游算子。
    • reduceGroup主要功能是通過 buildLabelIndexLabelArray 去重 “labels名字”，然後給每一個label一個ID，得到一個 <labels, ID>的map，最後返回是二元組(map, labels)，即({prefix1=0, prefix0=1},[prefix1, prefix0])。從後文看，<labels, ID>Map看來是多分類才用到。二分類只用到了labels。
    • mapPartition 分區調用 CalLabelDetailLocal 來計算混淆矩陣，主要是分區調用getDetailStatistics，前文中得到的二元組(map, labels)會作為參數傳遞進來 。
      • getDetailStatistics 遍歷 rows 數據，提取每一個item（比如 “prefix1,{“prefix1”: 0.8, “prefix0”: 0.2}”），然後通過updateBinaryMetricsSummary累積計算混淆矩陣所需數據。
        • updateBinaryMetricsSummary 把 [0,1] 分成假設 100000個桶(bin)。所以得到positiveBin / negativeBin 兩個100000的數組。positiveBin就是 TP + FP，negativeBin就是 TN + FN。
          • 如果某個 sample 為 正例 (positive value) 的概率是 p, 則該 sample 對應的 bin index 就是 p * 100000。如果 p 被預測為正例 (positive value) ，則positiveBin[index]++，
          • 否則就是被預測為負例(negative value) ，則negativeBin[index]++。
  • 綜合reduce上述計算結果，metrics = res.reduce(new EvaluationUtil.ReduceBaseMetrics());
    • 具體計算是在BinaryMetricsSummary.merge，其作用就是Merge the bins, and add the logLoss。
  • 把最終數值輸入到 output table，setOutput(metrics.flatMap(new EvaluationUtil.SaveDataAsParams()..);
    • 歸併所有BaseMetrics后，得到total BaseMetrics，計算indexes存入params。collector.collect(t.toMetrics().serialize());
      • 實際業務在BinaryMetricsSummary.toMetrics，即基於bin的信息計算，然後存儲到params。
        • extractMatrixThreCurve函數取出非空的bins，據此計算出ConfusionMatrix array（混淆矩陣）, threshold array, rocCurve/recallPrecisionCurve/LiftChart.
          • 遍歷bins中每一個有意義的點，計算出totalTrue和totalFalse，並且在每一個點上計算：
          • curTrue += positiveBin[index]; curFalse += negativeBin[index];
          • 得到該點的混淆矩陣 new ConfusionMatrix(new long[][] {{curTrue, curFalse}, {totalTrue – curTrue, totalFalse – curFalse}});
          • 得到 tpr = (totalTrue == 0 ? 1.0 : 1.0 * curTrue / totalTrue);
          • rocCurve，recallPrecisionCurve，liftChart在該點對應的數據；
        • 依據曲線內容計算並且存儲 AUC/PRC/KS
        • 對生成的rocCurve/recallPrecisionCurve/LiftChart輸出進行抽樣
        • 依據抽樣后的輸出存儲 RocCurve/RecallPrecisionCurve/LiftChar
        • 存儲正例樣本的度量指標
        • 存儲Logloss
        • Pick the middle point where threshold is 0.5.

3.2.1 calLabelPredDetailLocal

本函數按照partition分別計算評估指標 evaluation metrics。是的，這代碼很短，但是有個地方需要注意。有時候越簡單的地方越容易疏漏。容易疏漏點是：

第一行代碼的結果 labels 是第二行代碼的參數，而並非第二行主體。第二行代碼主體和第一行代碼主體一樣，都是data。

private static DataSet<BaseMetricsSummary> calLabelPredDetailLocal(DataSet<Row> data, final String positiveValue, oolean binary) {
  
    DataSet<Tuple2<Map<String, Integer>, String[]>> labels = data.flatMap(new FlatMapFunction<Row, String>() {
        @Override
        public void flatMap(Row row, Collector<String> collector) {
            TreeMap<String, Double> labelProbMap;
            if (EvaluationUtil.checkRowFieldNotNull(row)) {
                labelProbMap = EvaluationUtil.extractLabelProbMap(row);
                labelProbMap.keySet().forEach(collector::collect);
                collector.collect(row.getField(0).toString());
            }
        }
    }).reduceGroup(new EvaluationUtil.DistinctLabelIndexMap(binary, positiveValue));

    return data
        .rebalance()
        .mapPartition(new CalLabelDetailLocal(binary))
        .withBroadcastSet(labels, LABELS);
}

calLabelPredDetailLocal中具體分為三步驟：

  • 在flatMap會從label列和prediction列中，取出所有labels（注意是取出labels的名字 ），發送給下游算子。
  • reduceGroup的主要功能是去重 “labels名字”，然後給每一個label一個ID，最後結果是一個<labels, ID>Map。
  • mapPartition 是分區調用 CalLabelDetailLocal 來計算混淆矩陣。

下面具體看看。

3.2.1.1 flatMap

在flatMap中，主要是從label列和prediction列中，取出所有labels（注意是取出labels的名字 ），發送給下游算子。

EvaluationUtil.extractLabelProbMap 作用就是解析輸入的json，獲得具體detailInput中的信息。

下游算子是reduceGroup，所以Flink runtime會對這些labels自動去重。如果對這部分有興趣，可以參見我之前介紹reduce的文章。CSDN ： [源碼解析] Flink的groupBy和reduce究竟做了什麼 博客園 : [源碼解析] Flink的groupBy和reduce究竟做了什麼

程序中變量如下

row = {Row@8922} "prefix1,{"prefix1": 0.9, "prefix0": 0.1}"
 fields = {Object[2]@8925} 
  0 = "prefix1"
  1 = "{"prefix1": 0.9, "prefix0": 0.1}"
    
labelProbMap = {TreeMap@9008}  size = 2
 "prefix0" -> {Double@9015} 0.1
 "prefix1" -> {Double@9017} 0.9
    
labelProbMap.keySet().forEach(collector::collect); //這裏發送 "prefix0", "prefix1" 
collector.collect(row.getField(0).toString());  // 這裏發送 "prefix1"   
// 因為下一個操作是reduceGroup，所以這些label會被runtime去重
3.2.1.2 reduceGroup

主要功能是通過buildLabelIndexLabelArray去重labels，然後給每一個label一個ID，最後結果是一個<labels, ID>的Map。

reduceGroup(new EvaluationUtil.DistinctLabelIndexMap(binary, positiveValue));

DistinctLabelIndexMap的作用是從label列和prediction列中，取出所有不同的labels，返回一個<labels, ID>的map，根據後續代碼看，這個map是多分類才用到。Get all the distinct labels from label column and prediction column, and return the map of labels and their IDs.

前面已經提到，這裏的參數rows已經被自動去重。

public static class DistinctLabelIndexMap implements
    GroupReduceFunction<String, Tuple2<Map<String, Integer>, String[]>> {
    ......
    @Override
    public void reduce(Iterable<String> rows, Collector<Tuple2<Map<String, Integer>, String[]>> collector) throws Exception {
        HashSet<String> labels = new HashSet<>();
        rows.forEach(labels::add);
        collector.collect(buildLabelIndexLabelArray(labels, binary, positiveValue));
    }
}

// 變量為
labels = {HashSet@9008}  size = 2
 0 = "prefix1"
 1 = "prefix0"
binary = true

buildLabelIndexLabelArray的作用是給每一個label一個ID，得到一個 <labels, ID>的map，最後返回是二元組(map, labels)，即({prefix1=0, prefix0=1},[prefix1, prefix0])。

// Give each label an ID, return a map of label and ID.
public static Tuple2<Map<String, Integer>, String[]> buildLabelIndexLabelArray(HashSet<String> set,boolean binary, String positiveValue) {
    String[] labels = set.toArray(new String[0]);
    Arrays.sort(labels, Collections.reverseOrder());

    Map<String, Integer> map = new HashMap<>(labels.length);
    if (binary && null != positiveValue) {
        if (labels[1].equals(positiveValue)) {
            labels[1] = labels[0];
            labels[0] = positiveValue;
        } 
        map.put(labels[0], 0);
        map.put(labels[1], 1);
    } else {
        for (int i = 0; i < labels.length; i++) {
            map.put(labels[i], i);
        }
    }
    return Tuple2.of(map, labels);
}

// 程序變量如下
labels = {String[2]@9013} 
 0 = "prefix1"
 1 = "prefix0"
map = {HashMap@9014}  size = 2
 "prefix1" -> {Integer@9020} 0
 "prefix0" -> {Integer@9021} 1
3.2.1.3 mapPartition

這裏主要功能是分區調用 CalLabelDetailLocal 來為後來計算混淆矩陣做準備。

return data
    .rebalance()
    .mapPartition(new CalLabelDetailLocal(binary)) //這裡是業務所在
    .withBroadcastSet(labels, LABELS);

具體工作是 CalLabelDetailLocal 完成的，其作用是分區調用getDetailStatistics

// Calculate the confusion matrix based on the label and predResult.
static class CalLabelDetailLocal extends RichMapPartitionFunction<Row, BaseMetricsSummary> {
        private Tuple2<Map<String, Integer>, String[]> map;
        private boolean binary;

        @Override
        public void open(Configuration parameters) throws Exception {
            List<Tuple2<Map<String, Integer>, String[]>> list = getRuntimeContext().getBroadcastVariable(LABELS);
            this.map = list.get(0);// 前文生成的二元組(map, labels)
        }

        @Override
        public void mapPartition(Iterable<Row> rows, Collector<BaseMetricsSummary> collector) {
            // 調用到了 getDetailStatistics
            collector.collect(getDetailStatistics(rows, binary, map));
        }
    }

getDetailStatistics 的作用是：初始化分類評估的度量指標 base classification evaluation metrics，累積計算混淆矩陣需要的數據。主要就是遍歷 rows 數據，提取每一個item（比如 “prefix1,{“prefix1”: 0.8, “prefix0”: 0.2}”），然後累積計算混淆矩陣所需數據。

// Initialize the base classification evaluation metrics. There are two cases: BinaryClassMetrics and MultiClassMetrics.
    private static BaseMetricsSummary getDetailStatistics(Iterable<Row> rows,
                                         String positiveValue,
                                         boolean binary,
                                         Tuple2<Map<String, Integer>, String[]> tuple) {
        BinaryMetricsSummary binaryMetricsSummary = null;
        MultiMetricsSummary multiMetricsSummary = null;
        Tuple2<Map<String, Integer>, String[]> labelIndexLabelArray = tuple;  // 前文生成的二元組(map, labels)

        Iterator<Row> iterator = rows.iterator();
        Row row = null;
        while (iterator.hasNext() && !checkRowFieldNotNull(row)) {
            row = iterator.next();
        }

        Map<String, Integer> labelIndexMap = null;
        if (binary) {
           // 二分法在這裏 
            binaryMetricsSummary = new BinaryMetricsSummary(
                new long[ClassificationEvaluationUtil.DETAIL_BIN_NUMBER],
                new long[ClassificationEvaluationUtil.DETAIL_BIN_NUMBER],
                labelIndexLabelArray.f1, 0.0, 0L);
        } else {
            // 
            labelIndexMap = labelIndexLabelArray.f0; // 前文生成的<labels, ID>Map看來是多分類才用到。
            multiMetricsSummary = new MultiMetricsSummary(
                new long[labelIndexMap.size()][labelIndexMap.size()],
                labelIndexLabelArray.f1, 0.0, 0L);
        }

        while (null != row) {
            if (checkRowFieldNotNull(row)) {
                TreeMap<String, Double> labelProbMap = extractLabelProbMap(row);
                String label = row.getField(0).toString();
                if (ArrayUtils.indexOf(labelIndexLabelArray.f1, label) >= 0) {
                    if (binary) {
                        // 二分法在這裏 
                        updateBinaryMetricsSummary(labelProbMap, label, binaryMetricsSummary);
                    } else {
                        updateMultiMetricsSummary(labelProbMap, label, labelIndexMap, multiMetricsSummary);
                    }
                }
            }
            row = iterator.hasNext() ? iterator.next() : null;
        }

        return binary ? binaryMetricsSummary : multiMetricsSummary;
}

//變量如下
tuple = {Tuple2@9252} "({prefix1=0, prefix0=1},[prefix1, prefix0])"
 f0 = {HashMap@9257}  size = 2
  "prefix1" -> {Integer@9264} 0
  "prefix0" -> {Integer@9266} 1
 f1 = {String[2]@9258} 
  0 = "prefix1"
  1 = "prefix0"
 
row = {Row@9271} "prefix1,{"prefix1": 0.8, "prefix0": 0.2}"
 fields = {Object[2]@9276} 
  0 = "prefix1"
  1 = "{"prefix1": 0.8, "prefix0": 0.2}"
    
labelIndexLabelArray = {Tuple2@9240} "({prefix1=0, prefix0=1},[prefix1, prefix0])"
 f0 = {HashMap@9288}  size = 2
  "prefix1" -> {Integer@9294} 0
  "prefix0" -> {Integer@9296} 1
 f1 = {String[2]@9242} 
  0 = "prefix1"
  1 = "prefix0"
    
labelProbMap = {TreeMap@9342}  size = 2
 "prefix0" -> {Double@9378} 0.1
 "prefix1" -> {Double@9380} 0.9

先回憶下混淆矩陣：

預測值 0 預測值 1
真實值 0 TN FP
真實值 1 FN TP

針對混淆矩陣，BinaryMetricsSummary 的作用是Save the evaluation data for binary classification。函數具體計算思路是：

  • 把 [0,1] 分成ClassificationEvaluationUtil.DETAIL_BIN_NUMBER（100000）這麼多桶(bin)。所以binaryMetricsSummary的positiveBin/negativeBin分別是兩個100000的數組。如果某一個 sample 為 正例(positive value) 的概率是 p, 則該 sample 對應的 bin index 就是 p * 100000。如果 p 被預測為正例(positive value) ，則positiveBin[index]++，否則就是被預測為負例(negative value) ，則negativeBin[index]++。positiveBin就是 TP + FP，negativeBin就是 TN + FN。

  • 所以這裡會遍歷輸入，如果某一個輸入（以"prefix1", "{\"prefix1\": 0.9, \"prefix0\": 0.1}"為例），0.9 是prefix1(正例) 的概率，0.1 是為prefix0(負例) 的概率。

    • 既然這個算法選擇了 prefix1(正例) ，所以就說明此算法是判別成 positive 的，所以在 positiveBin 的 90000 處 + 1。
    • 假設這個算法選擇了 prefix0(負例) ，則說明此算法是判別成 negative 的，所以應該在 negativeBin 的 90000 處 + 1。

具體對應我們示例代碼的5個採樣，分類如下：

Row.of("prefix1", "{\"prefix1\": 0.9, \"prefix0\": 0.1}"),  positiveBin 90000處+1
Row.of("prefix1", "{\"prefix1\": 0.8, \"prefix0\": 0.2}"),  positiveBin 80000處+1
Row.of("prefix1", "{\"prefix1\": 0.7, \"prefix0\": 0.3}"),  positiveBin 70000處+1
Row.of("prefix0", "{\"prefix1\": 0.75, \"prefix0\": 0.25}"), negativeBin 75000處+1
Row.of("prefix0", "{\"prefix1\": 0.6, \"prefix0\": 0.4}")  negativeBin 60000處+1

具體代碼如下

public static void updateBinaryMetricsSummary(TreeMap<String, Double> labelProbMap,
                                              String label,
                                              BinaryMetricsSummary binaryMetricsSummary) {
    binaryMetricsSummary.total++;
    binaryMetricsSummary.logLoss += extractLogloss(labelProbMap, label);

    double d = labelProbMap.get(binaryMetricsSummary.labels[0]);
    int idx = d == 1.0 ? ClassificationEvaluationUtil.DETAIL_BIN_NUMBER - 1 :
        (int)Math.floor(d * ClassificationEvaluationUtil.DETAIL_BIN_NUMBER);
    if (idx >= 0 && idx < ClassificationEvaluationUtil.DETAIL_BIN_NUMBER) {
        if (label.equals(binaryMetricsSummary.labels[0])) {
            binaryMetricsSummary.positiveBin[idx] += 1;
        } else if (label.equals(binaryMetricsSummary.labels[1])) {
            binaryMetricsSummary.negativeBin[idx] += 1;
        } else {
					.....
        }
    }
}

private static double extractLogloss(TreeMap<String, Double> labelProbMap, String label) {
   Double prob = labelProbMap.get(label);
   prob = null == prob ? 0. : prob;
   return -Math.log(Math.max(Math.min(prob, 1 - LOG_LOSS_EPS), LOG_LOSS_EPS));
}

// 變量如下
ClassificationEvaluationUtil.DETAIL_BIN_NUMBER=100000
  
// 當 "prefix1", "{\"prefix1\": 0.9, \"prefix0\": 0.1}" 時候
labelProbMap = {TreeMap@9305}  size = 2
 "prefix0" -> {Double@9331} 0.1
 "prefix1" -> {Double@9333} 0.9
  
d = 0.9
idx = 90000
binaryMetricsSummary = {BinaryMetricsSummary@9262} 
 labels = {String[2]@9242} 
  0 = "prefix1"
  1 = "prefix0"
 total = 1
 positiveBin = {long[100000]@9263}  // 90000處+1
 negativeBin = {long[100000]@9264} 
 logLoss = 0.10536051565782628
   
// 當 "prefix0", "{\"prefix1\": 0.6, \"prefix0\": 0.4}" 時候  
labelProbMap = {TreeMap@9514}  size = 2
 "prefix0" -> {Double@9546} 0.4
 "prefix1" -> {Double@9547} 0.6
   
d = 0.6
idx = 60000    
 binaryMetricsSummary = {BinaryMetricsSummary@9262} 
 labels = {String[2]@9242} 
  0 = "prefix1"
  1 = "prefix0"
 total = 2
 positiveBin = {long[100000]@9263}  
 negativeBin = {long[100000]@9264} // 60000處+1
 logLoss = 1.0216512475319812

3.2.2 ReduceBaseMetrics

ReduceBaseMetrics作用是把局部計算的 BaseMetrics 聚合起來。

DataSet<BaseMetricsSummary> metrics = res
    .reduce(new EvaluationUtil.ReduceBaseMetrics());

ReduceBaseMetrics如下

public static class ReduceBaseMetrics implements ReduceFunction<BaseMetricsSummary> {
    @Override
    public BaseMetricsSummary reduce(BaseMetricsSummary t1, BaseMetricsSummary t2) throws Exception {
        return null == t1 ? t2 : t1.merge(t2);
    }
}

具體計算是在BinaryMetricsSummary.merge，其作用就是Merge the bins, and add the logLoss。

@Override
public BinaryMetricsSummary merge(BinaryMetricsSummary binaryClassMetrics) {
    for (int i = 0; i < this.positiveBin.length; i++) {
        this.positiveBin[i] += binaryClassMetrics.positiveBin[i];
    }
    for (int i = 0; i < this.negativeBin.length; i++) {
        this.negativeBin[i] += binaryClassMetrics.negativeBin[i];
    }
    this.logLoss += binaryClassMetrics.logLoss;
    this.total += binaryClassMetrics.total;
    return this;
}

// 程序變量是
this = {BinaryMetricsSummary@9316} 
 labels = {String[2]@9322} 
  0 = "prefix1"
  1 = "prefix0"
 total = 2
 positiveBin = {long[100000]@9320} 
 negativeBin = {long[100000]@9323} 
 logLoss = 1.742969305058623

3.2.3 SaveDataAsParams

this.setOutput(metrics.flatMap(new EvaluationUtil.SaveDataAsParams()),
    new String[] {DATA_OUTPUT}, new TypeInformation[] {Types.STRING});

當歸併所有BaseMetrics之後，得到了total BaseMetrics，計算indexes，存入到params。

public static class SaveDataAsParams implements FlatMapFunction<BaseMetricsSummary, Row> {
    @Override
    public void flatMap(BaseMetricsSummary t, Collector<Row> collector) throws Exception {
        collector.collect(t.toMetrics().serialize());
    }
}

實際業務在BinaryMetricsSummary.toMetrics中完成，即基於bin的信息計算，得到confusionMatrix array, threshold array, rocCurve/recallPrecisionCurve/LiftChart等等，然後存儲到params。

public BinaryClassMetrics toMetrics() {
    Params params = new Params();
    // 生成若干曲線，比如rocCurve/recallPrecisionCurve/LiftChart
    Tuple3<ConfusionMatrix[], double[], EvaluationCurve[]> matrixThreCurve =
        extractMatrixThreCurve(positiveBin, negativeBin, total);

    // 依據曲線內容計算並且存儲 AUC/PRC/KS
    setCurveAreaParams(params, matrixThreCurve.f2);

    // 對生成的rocCurve/recallPrecisionCurve/LiftChart輸出進行抽樣
    Tuple3<ConfusionMatrix[], double[], EvaluationCurve[]> sampledMatrixThreCurve = sample(
        PROBABILITY_INTERVAL, matrixThreCurve);

    // 依據抽樣后的輸出存儲 RocCurve/RecallPrecisionCurve/LiftChar
    setCurvePointsParams(params, sampledMatrixThreCurve);
    ConfusionMatrix[] matrices = sampledMatrixThreCurve.f0;
  
    // 存儲正例樣本的度量指標
    setComputationsArrayParams(params, sampledMatrixThreCurve.f1, sampledMatrixThreCurve.f0);
  
    // 存儲Logloss
    setLoglossParams(params, logLoss, total);
  
    // Pick the middle point where threshold is 0.5.
    int middleIndex = getMiddleThresholdIndex(sampledMatrixThreCurve.f1);  
    setMiddleThreParams(params, matrices[middleIndex], labels);
    return new BinaryClassMetrics(params);
}

extractMatrixThreCurve是全文重點。這裡是 Extract the bins who are not empty, keep the middle threshold 0.5，然後初始化了 RocCurve, Recall-Precision Curve and Lift Curve，計算出ConfusionMatrix array（混淆矩陣）, threshold array, rocCurve/recallPrecisionCurve/LiftChart.。

/**
 * Extract the bins who are not empty, keep the middle threshold 0.5.
 * Initialize the RocCurve, Recall-Precision Curve and Lift Curve.
 * RocCurve: (FPR, TPR), starts with (0,0). Recall-Precision Curve: (recall, precision), starts with (0, p), p is the precision with the lowest. LiftChart: (TP+FP/total, TP), starts with (0,0). confusion matrix = [TP FP][FN * TN].
 *
 * @param positiveBin positiveBins.
 * @param negativeBin negativeBins.
 * @param total       sample number
 * @return ConfusionMatrix array, threshold array, rocCurve/recallPrecisionCurve/LiftChart.
 */
static Tuple3<ConfusionMatrix[], double[], EvaluationCurve[]> extractMatrixThreCurve(long[] positiveBin, long[] negativeBin, long total) {
    ArrayList<Integer> effectiveIndices = new ArrayList<>();
    long totalTrue = 0, totalFalse = 0;
  
    // 計算totalTrue，totalFalse，effectiveIndices
    for (int i = 0; i < ClassificationEvaluationUtil.DETAIL_BIN_NUMBER; i++) {
        if (0L != positiveBin[i] || 0L != negativeBin[i]
            || i == ClassificationEvaluationUtil.DETAIL_BIN_NUMBER / 2) {
            effectiveIndices.add(i);
            totalTrue += positiveBin[i];
            totalFalse += negativeBin[i];
        }
    }

// 以我們例子，得到  
effectiveIndices = {ArrayList@9273}  size = 6
 0 = {Integer@9277} 50000 //這裏加入了中間點
 1 = {Integer@9278} 60000
 2 = {Integer@9279} 70000
 3 = {Integer@9280} 75000
 4 = {Integer@9281} 80000
 5 = {Integer@9282} 90000
totalTrue = 3
totalFalse = 2
  
    // 繼續初始化，生成若干curve
    final int length = effectiveIndices.size();
    final int newLen = length + 1;
    final double m = 1.0 / ClassificationEvaluationUtil.DETAIL_BIN_NUMBER;
    EvaluationCurvePoint[] rocCurve = new EvaluationCurvePoint[newLen];
    EvaluationCurvePoint[] recallPrecisionCurve = new EvaluationCurvePoint[newLen];
    EvaluationCurvePoint[] liftChart = new EvaluationCurvePoint[newLen];
    ConfusionMatrix[] data = new ConfusionMatrix[newLen];
    double[] threshold = new double[newLen];
    long curTrue = 0;
    long curFalse = 0;
  
// 以我們例子，得到 
length = 6
newLen = 7
m = 1.0E-5
  
    // 計算, 其中rocCurve，recallPrecisionCurve，liftChart 都可以從代碼中看出
    for (int i = 1; i < newLen; i++) {
        int index = effectiveIndices.get(length - i);
        curTrue += positiveBin[index];
        curFalse += negativeBin[index];
        threshold[i] = index * m;
        // 計算出混淆矩陣
        data[i] = new ConfusionMatrix(
            new long[][] {{curTrue, curFalse}, {totalTrue - curTrue, totalFalse - curFalse}});
        double tpr = (totalTrue == 0 ? 1.0 : 1.0 * curTrue / totalTrue);
        // 比如當 90000 這點，得到 curTrue = 1 curFalse = 0 i = 1 index = 90000 tpr = 0.3333333333333333。totalTrue = 3 totalFalse = 2， 
        // 我們也知道，TPR = TP / (TP + FN) ，所以可以計算 tpr = 1 / 3   
        rocCurve[i] = new EvaluationCurvePoint(totalFalse == 0 ? 1.0 : 1.0 * curFalse / totalFalse, tpr, threshold[i]);
        recallPrecisionCurve[i] = new EvaluationCurvePoint(tpr, curTrue + curTrue == 0 ? 1.0 : 1.0 * curTrue / (curTrue + curFalse), threshold[i]);
        liftChart[i] = new EvaluationCurvePoint(1.0 * (curTrue + curFalse) / total, curTrue, threshold[i]);
    }
  
// 以我們例子，得到 
curTrue = 3
curFalse = 2
  
threshold = {double[7]@9349} 
 0 = 0.0
 1 = 0.9
 2 = 0.8
 3 = 0.7500000000000001
 4 = 0.7000000000000001
 5 = 0.6000000000000001
 6 = 0.5  
   
rocCurve = {EvaluationCurvePoint[7]@9315} 
 1 = {EvaluationCurvePoint@9440} 
  x = 0.0
  y = 0.3333333333333333
  p = 0.9
 2 = {EvaluationCurvePoint@9448} 
  x = 0.0
  y = 0.6666666666666666
  p = 0.8
 3 = {EvaluationCurvePoint@9449} 
  x = 0.5
  y = 0.6666666666666666
  p = 0.7500000000000001
 4 = {EvaluationCurvePoint@9450} 
  x = 0.5
  y = 1.0
  p = 0.7000000000000001
 5 = {EvaluationCurvePoint@9451} 
  x = 1.0
  y = 1.0
  p = 0.6000000000000001
 6 = {EvaluationCurvePoint@9452} 
  x = 1.0
  y = 1.0
  p = 0.5
    
recallPrecisionCurve = {EvaluationCurvePoint[7]@9320} 
 1 = {EvaluationCurvePoint@9444} 
  x = 0.3333333333333333
  y = 1.0
  p = 0.9
 2 = {EvaluationCurvePoint@9453} 
  x = 0.6666666666666666
  y = 1.0
  p = 0.8
 3 = {EvaluationCurvePoint@9454} 
  x = 0.6666666666666666
  y = 0.6666666666666666
  p = 0.7500000000000001
 4 = {EvaluationCurvePoint@9455} 
  x = 1.0
  y = 0.75
  p = 0.7000000000000001
 5 = {EvaluationCurvePoint@9456} 
  x = 1.0
  y = 0.6
  p = 0.6000000000000001
 6 = {EvaluationCurvePoint@9457} 
  x = 1.0
  y = 0.6
  p = 0.5
    
liftChart = {EvaluationCurvePoint[7]@9325} 
 1 = {EvaluationCurvePoint@9458} 
  x = 0.2
  y = 1.0
  p = 0.9
 2 = {EvaluationCurvePoint@9459} 
  x = 0.4
  y = 2.0
  p = 0.8
 3 = {EvaluationCurvePoint@9460} 
  x = 0.6
  y = 2.0
  p = 0.7500000000000001
 4 = {EvaluationCurvePoint@9461} 
  x = 0.8
  y = 3.0
  p = 0.7000000000000001
 5 = {EvaluationCurvePoint@9462} 
  x = 1.0
  y = 3.0
  p = 0.6000000000000001
 6 = {EvaluationCurvePoint@9463} 
  x = 1.0
  y = 3.0
  p = 0.5
    
data = {ConfusionMatrix[7]@9339} 
 0 = {ConfusionMatrix@9486} 
  longMatrix = {LongMatrix@9488} 
   matrix = {long[2][]@9491} 
    0 = {long[2]@9492} 
     0 = 0
     1 = 0
    1 = {long[2]@9493} 
     0 = 3
     1 = 2
   rowNum = 2
   colNum = 2
  labelCnt = 2
  total = 5
  actualLabelFrequency = {long[2]@9489} 
   0 = 3
   1 = 2
  predictLabelFrequency = {long[2]@9490} 
   0 = 0
   1 = 5
  tpCount = 2.0
  tnCount = 2.0
  fpCount = 3.0
  fnCount = 3.0
 1 = {ConfusionMatrix@9435} 
  longMatrix = {LongMatrix@9469} 
   matrix = {long[2][]@9472} 
    0 = {long[2]@9474} 
     0 = 1
     1 = 0
    1 = {long[2]@9475} 
     0 = 2
     1 = 2
   rowNum = 2
   colNum = 2
  labelCnt = 2
  total = 5
  actualLabelFrequency = {long[2]@9470} 
   0 = 3
   1 = 2
  predictLabelFrequency = {long[2]@9471} 
   0 = 1
   1 = 4
  tpCount = 3.0
  tnCount = 3.0
  fpCount = 2.0
  fnCount = 2.0
  ......  
    
    threshold[0] = 1.0;
    data[0] = new ConfusionMatrix(new long[][] {{0, 0}, {totalTrue, totalFalse}});
    rocCurve[0] = new EvaluationCurvePoint(0, 0, threshold[0]);
    recallPrecisionCurve[0] = new EvaluationCurvePoint(0, recallPrecisionCurve[1].getY(), threshold[0]);
    liftChart[0] = new EvaluationCurvePoint(0, 0, threshold[0]);

    return Tuple3.of(data, threshold, new EvaluationCurve[] {new EvaluationCurve(rocCurve),
        new EvaluationCurve(recallPrecisionCurve), new EvaluationCurve(liftChart)});
}

3.2.4 計算混淆矩陣

這裏再給大家講講混淆矩陣如何計算，這裏思路比較繞。

3.2.4.1 原始矩陣

調用之處是：

// 調用之處
data[i] = new ConfusionMatrix(
        new long[][] {{curTrue, curFalse}, {totalTrue - curTrue, totalFalse - curFalse}});
// 調用時候各種賦值
i = 1
index = 90000
totalTrue = 3
totalFalse = 2
curTrue = 1
curFalse = 0

得到原始矩陣，以下都有cur，說明只針對當前點來說。
curTrue = 1 curFalse = 0
totalTrue – curTrue = 2 totalFalse – curFalse = 2
3.2.4.2 計算標籤

後續ConfusionMatrix計算中，由此可以得到

actualLabelFrequency = longMatrix.getColSums();
predictLabelFrequency = longMatrix.getRowSums();

actualLabelFrequency = {long[2]@9322} 
 0 = 3
 1 = 2
predictLabelFrequency = {long[2]@9323} 
 0 = 1
 1 = 4

可以看出來，Alink算法認為：每列的sum和實際標籤有關；每行sum和預測標籤有關。

得到新矩陣如下

predictLabelFrequency
curTrue = 1 curFalse = 0 1 = curTrue + curFalse
totalTrue – curTrue = 2 totalFalse – curFalse = 2 4 = total – curTrue – curFalse
actualLabelFrequency 3 = totalTrue 2 = totalFalse

後續計算將要基於這些來計算：

計算中就用到longMatrix 對角線上的數據，即longMatrix(0)(0)和 longMatrix(1)(1)。一定要注意，這裏考慮的都是 當前狀態 (畫重點強調)。

longMatrix(0)(0) ：curTrue

longMatrix(1)(1) ：totalFalse – curFalse

totalFalse ：( TN + FN )

totalTrue ：( TP + FP )

double numTrueNegative(Integer labelIndex) {
  // labelIndex為 0 時候，return 1 + 5 - 1 - 3 = 2;
  // labelIndex為 1 時候，return 2 + 5 - 4 - 2 = 1;
	return null == labelIndex ? tnCount : longMatrix.getValue(labelIndex, labelIndex) + total - predictLabelFrequency[labelIndex] - actualLabelFrequency[labelIndex];
}

double numTruePositive(Integer labelIndex) {
  // labelIndex為 0 時候，return 1; 這個是 curTrue，就是真實標籤是True，判別也是True。是TP
  // labelIndex為 1 時候，return 2; 這個是 totalFalse - curFalse，總判別錯 - 當前判別錯。這就意味着“本來判別錯了但是當前沒有發現”，所以認為在當前狀態下，這也算是TP
	return null == labelIndex ? tpCount : longMatrix.getValue(labelIndex, labelIndex);
}

double numFalseNegative(Integer labelIndex) {
  // labelIndex為 0 時候，return 3 - 1; 
  // actualLabelFrequency[0] = totalTrue。所以return totalTrue - curTrue，即當前“全部正確”中沒有“判別為正確”，這個就可以認為是“判別錯了且判別為負”
  // labelIndex為 1 時候，return 2 - 2;   
  // actualLabelFrequency[1] = totalFalse。所以return totalFalse - ( totalFalse - curFalse )  = curFalse
	return null == labelIndex ? fnCount : actualLabelFrequency[labelIndex] - longMatrix.getValue(labelIndex, labelIndex);
}

double numFalsePositive(Integer labelIndex) {
  // labelIndex為 0 時候，return 1 - 1;
  // predictLabelFrequency[0] = curTrue + curFalse。
  // 所以 return = curTrue + curFalse - curTrue = curFalse = current( TN + FN ) 這可以認為是判斷錯了實際是正確標籤
  // labelIndex為 1 時候，return 4 - 2; 
  // predictLabelFrequency[1] = total - curTrue - curFalse。
  // 所以 return = total - curTrue - curFalse - (totalFalse - curFalse) = totalTrue - curTrue = ( TP + FP ) - currentTP = currentFP 
	return null == labelIndex ? fpCount : predictLabelFrequency[labelIndex] - longMatrix.getValue(labelIndex, labelIndex);
}

// 最後得到
tpCount = 3.0
tnCount = 3.0
fpCount = 2.0
fnCount = 2.0
3.2.4.3 具體代碼
// 具體計算 
public ConfusionMatrix(LongMatrix longMatrix) {
  
longMatrix = {LongMatrix@9297} 
  0 = {long[2]@9324} 
   0 = 1
   1 = 0
  1 = {long[2]@9325} 
   0 = 2
   1 = 2
     
    this.longMatrix = longMatrix;
    labelCnt = this.longMatrix.getRowNum();
    // 這裏就是計算
    actualLabelFrequency = longMatrix.getColSums();
    predictLabelFrequency = longMatrix.getRowSums();
  
actualLabelFrequency = {long[2]@9322} 
 0 = 3
 1 = 2
predictLabelFrequency = {long[2]@9323} 
 0 = 1
 1 = 4  
labelCnt = 2
total = 5  

    total = longMatrix.getTotal();
    for (int i = 0; i < labelCnt; i++) {
        tnCount += numTrueNegative(i);
        tpCount += numTruePositive(i);
        fnCount += numFalseNegative(i);
        fpCount += numFalsePositive(i);
    }
}

0x04 流處理

4.1 示例

Alink原有python示例代碼中，Stream部分是沒有輸出的，因為MemSourceStreamOp沒有和時間相關聯，而Alink中沒有提供基於時間的StreamOperator，所以只能自己仿照MemSourceBatchOp寫了一個。雖然代碼有些丑，但是至少可以提供輸出，這樣就能夠調試。

4.1.1 主類

public class EvalBinaryClassExampleStream {

    AlgoOperator getData(boolean isBatch) {
        Row[] rows = new Row[]{
                Row.of("prefix1", "{\"prefix1\": 0.9, \"prefix0\": 0.1}")
        };
        String[] schema = new String[]{"label", "detailInput"};
        if (isBatch) {
            return new MemSourceBatchOp(rows, schema);
        } else {
            return new TimeMemSourceStreamOp(rows, schema, new EvalBinaryStreamSource());
        }
    }

    public static void main(String[] args) throws Exception {
        EvalBinaryClassExampleStream test = new EvalBinaryClassExampleStream();
        StreamOperator streamData = (StreamOperator) test.getData(false);
        StreamOperator sOp = new EvalBinaryClassStreamOp()
                .setLabelCol("label")
                .setPredictionDetailCol("detailInput")
                .setTimeInterval(1)
                .linkFrom(streamData);
        sOp.print();
        StreamOperator.execute();
    }
}

4.1.2 TimeMemSourceStreamOp

這個是我自己炮製的。借鑒了MemSourceStreamOp。

public final class TimeMemSourceStreamOp extends StreamOperator<TimeMemSourceStreamOp> {

    public TimeMemSourceStreamOp(Row[] rows, String[] colNames, EvalBinaryStrSource source) {
        super(null);
        init(source, Arrays.asList(rows), colNames);
    }

    private void init(EvalBinaryStreamSource source, List <Row> rows, String[] colNames) {
        Row first = rows.iterator().next();
        int arity = first.getArity();
        TypeInformation <?>[] types = new TypeInformation[arity];

        for (int i = 0; i < arity; ++i) {
            types[i] = TypeExtractor.getForObject(first.getField(i));
        }

        init(source, colNames, types);
    }

    private void init(EvalBinaryStreamSource source, String[] colNames, TypeInformation <?>[] colTypes) {
        DataStream <Row> dastr = MLEnvironmentFactory.get(getMLEnvironmentId())
                .getStreamExecutionEnvironment().addSource(source);
        StringBuilder sbd = new StringBuilder();
        sbd.append(colNames[0]);
      
        for (int i = 1; i < colNames.length; i++) {
            sbd.append(",").append(colNames[i]);
        }
        this.setOutput(dastr, colNames, colTypes);
    }

    @Override
    public TimeMemSourceStreamOp linkFrom(StreamOperator<?>... inputs) {
        return null;
    }
}

4.1.3 Source

定時提供Row，加入了隨機數，讓概率有變化。

class EvalBinaryStreamSource extends RichSourceFunction[Row] {

  override def run(ctx: SourceFunction.SourceContext[Row]) = {
    while (true) {
      val rdm = Math.random() // 這裏加入了隨機數，讓概率有變化
      val rows: Array[Row] = Array[Row](
        Row.of("prefix1", "{\"prefix1\": " + rdm + ", \"prefix0\": " + (1-rdm) + "}"),
        Row.of("prefix1", "{\"prefix1\": 0.8, \"prefix0\": 0.2}"),
        Row.of("prefix1", "{\"prefix1\": 0.7, \"prefix0\": 0.3}"),
        Row.of("prefix0", "{\"prefix1\": 0.75, \"prefix0\": 0.25}"),
        Row.of("prefix0", "{\"prefix1\": 0.6, \"prefix0\": 0.4}"))
      for(row <- rows) {
        println(s"當前值：$row")
        ctx.collect(row)
      }
      Thread.sleep(1000)
    }
  }

  override def cancel() = ???
}

4.2 BaseEvalClassStreamOp

Alink流處理類是 EvalBinaryClassStreamOp，主要工作在其基類 BaseEvalClassStreamOp，所以我們重點看後者。

public class BaseEvalClassStreamOp<T extends BaseEvalClassStreamOp<T>> extends StreamOperator<T> {
    @Override
    public T linkFrom(StreamOperator<?>... inputs) {
        StreamOperator<?> in = checkAndGetFirst(inputs);
        String labelColName = this.get(MultiEvaluationStreamParams.LABEL_COL);
        String positiveValue = this.get(BinaryEvaluationStreamParams.POS_LABEL_VAL_STR);
        Integer timeInterval = this.get(MultiEvaluationStreamParams.TIME_INTERVAL);

        ClassificationEvaluationUtil.Type type = ClassificationEvaluationUtil.judgeEvaluationType(this.getParams());

        DataStream<BaseMetricsSummary> statistics;

        switch (type) {
            case PRED_RESULT: {
              ......
            }
            case PRED_DETAIL: {               
                String predDetailColName = this.get(MultiEvaluationStreamParams.PREDICTION_DETAIL_COL);
                // 
                PredDetailLabel eval = new PredDetailLabel(positiveValue, binary);
                // 獲取輸入數據，重點是timeWindowAll
                statistics = in.select(new String[] {labelColName, predDetailColName})
                    .getDataStream()
                    .timeWindowAll(Time.of(timeInterval, TimeUnit.SECONDS))
                    .apply(eval);
                break;
            }
        }
        // 把各個窗口的數據累積到 totalStatistics，注意，這裡是新變量了。
        DataStream<BaseMetricsSummary> totalStatistics = statistics
            .map(new EvaluationUtil.AllDataMerge())
            .setParallelism(1); // 并行度設置為1

        // 基於兩種 bins 計算&序列化，得到當前的 statistics
        DataStream<Row> windowOutput = statistics.map(
            new EvaluationUtil.SaveDataStream(ClassificationEvaluationUtil.WINDOW.f0));
        // 基於bins計算&序列化，得到累積的 totalStatistics
        DataStream<Row> allOutput = totalStatistics.map(
            new EvaluationUtil.SaveDataStream(ClassificationEvaluationUtil.ALL.f0));

      	// "當前" 和 "累積" 做聯合，最終返回
        DataStream<Row> union = windowOutput.union(allOutput);

        this.setOutput(union,
            new String[] {ClassificationEvaluationUtil.STATISTICS_OUTPUT, DATA_OUTPUT},
            new TypeInformation[] {Types.STRING, Types.STRING});

        return (T)this;
    }
}

具體業務是：

  • PredDetailLabel 會進行去重標籤名字 和 累積計算混淆矩陣所需數據
    • buildLabelIndexLabelArray 去重 “labels名字”，然後給每一個label一個ID，最後結果是一個<labels, ID>Map。
    • getDetailStatistics 遍歷 rows 數據，提取每一個item（比如 “prefix1,{“prefix1”: 0.8, “prefix0”: 0.2}”），然後通過updateBinaryMetricsSummary累積計算混淆矩陣所需數據。
  • 根據標籤從Window中獲取數據 statistics = in.select().getDataStream().timeWindowAll() .apply(eval);
  • EvaluationUtil.AllDataMerge 把各個窗口的數據累積到 totalStatistics 。
  • 得到windowOutput ——– EvaluationUtil.SaveDataStream，對”當前數據statistics”做處理。實際業務在BinaryMetricsSummary.toMetrics，即基於bin的信息計算，然後存儲到params，並序列化返回Row。
    • extractMatrixThreCurve函數取出非空的bins，據此計算出ConfusionMatrix array（混淆矩陣）, threshold array, rocCurve/recallPrecisionCurve/LiftChart.
    • 依據曲線內容計算並且存儲 AUC/PRC/KS
    • 對生成的rocCurve/recallPrecisionCurve/LiftChart輸出進行抽樣
    • 依據抽樣后的輸出存儲 RocCurve/RecallPrecisionCurve/LiftChar
    • 存儲正例樣本的度量指標
    • 存儲Logloss
    • Pick the middle point where threshold is 0.5.
  • 得到allOutput ——– EvaluationUtil.SaveDataStream , 對”累積數據totalStatistics”做處理。
    • 詳細處理流程同windowOutput。
  • windowOutput 和 allOutput 做聯合。最終返回 DataStream union = windowOutput.union(allOutput);

4.2.1 PredDetailLabel

static class PredDetailLabel implements AllWindowFunction<Row, BaseMetricsSummary, TimeWindow> {
    @Override
    public void apply(TimeWindow timeWindow, Iterable<Row> rows, Collector<BaseMetricsSummary> collector) throws Exception {
        HashSet<String> labels = new HashSet<>();
        // 首先還是獲取 labels 名字
        for (Row row : rows) {
            if (EvaluationUtil.checkRowFieldNotNull(row)) {
                labels.addAll(EvaluationUtil.extractLabelProbMap(row).keySet());
                labels.add(row.getField(0).toString());
            }
        }
labels = {HashSet@9757}  size = 2
 0 = "prefix1"
 1 = "prefix0"   
        // 之前介紹過，buildLabelIndexLabelArray 去重 "labels名字"，然後給每一個label一個ID，最後結果是一個<labels, ID>Map。
        // getDetailStatistics 遍歷 rows 數據，累積計算混淆矩陣所需數據（ "TP + FN"  /  "TN + FP"）。
        if (labels.size() > 0) {
            collector.collect(
                getDetailStatistics(rows, binary, buildLabelIndexLabelArray(labels, binary, positiveValue)));
        }
    }
}

4.2.2 AllDataMerge

EvaluationUtil.AllDataMerge 把各個窗口的數據累積

/**
 * Merge data from different windows.
 */
public static class AllDataMerge implements MapFunction<BaseMetricsSummary, BaseMetricsSummary> {
    private BaseMetricsSummary statistics;
    @Override
    public BaseMetricsSummary map(BaseMetricsSummary value) {
        this.statistics = (null == this.statistics ? value : this.statistics.merge(value));
        return this.statistics;
    }
}

4.2.3 SaveDataStream

SaveDataStream具體調用的函數之前批處理介紹過，實際業務在BinaryMetricsSummary.toMetrics，即基於bin的信息計算，存儲到params。

這裏與批處理不同的是直接就把”構建出的度量信息“返回給用戶。

public static class SaveDataStream implements MapFunction<BaseMetricsSummary, Row> {
    @Override
    public Row map(BaseMetricsSummary baseMetricsSummary) throws Exception {
        BaseMetricsSummary metrics = baseMetricsSummary;
        BaseMetrics baseMetrics = metrics.toMetrics();
        Row row = baseMetrics.serialize();
        return Row.of(funtionName, row.getField(0));
    }
}

// 最後得到的 row 其實就是最終返回給用戶的度量信息
row = {Row@10008} "{"PRC":"0.9164636268708667","SensitivityArray":"[0.38461538461538464,0.6923076923076923,0.6923076923076923,1.0,1.0,1.0]","ConfusionMatrix":"[[13,8],[0,0]]","MacroRecall":"0.5","MacroSpecificity":"0.5","FalsePositiveRateArray":"[0.0,0.0,0.5,0.5,1.0,1.0]" ...... 還有很多其他的

4.2.4 Union

DataStream<Row> windowOutput = statistics.map(
    new EvaluationUtil.SaveDataStream(ClassificationEvaluationUtil.WINDOW.f0));
DataStream<Row> allOutput = totalStatistics.map(
    new EvaluationUtil.SaveDataStream(ClassificationEvaluationUtil.ALL.f0));

DataStream<Row> union = windowOutput.union(allOutput);

最後返回兩種統計數據

4.2.4.1 allOutput
all|{"PRC":"0.7341146115890359","SensitivityArray":"[0.3333333333333333,0.3333333333333333,0.6666666666666666,0.7333333333333333,0.8,0.8,0.8666666666666667,0.8666666666666667,0.9333333333333333,1.0]","ConfusionMatrix":"[[13,10],[2,0]]","MacroRecall":"0.43333333333333335","MacroSpecificity":"0.43333333333333335","FalsePositiveRateArray":"[0.0,0.5,0.5,0.5,0.5,1.0,1.0,1.0,1.0,1.0]","TruePositiveRateArray":"[0.3333333333333333,0.3333333333333333,0.6666666666666666,0.7333333333333333,0.8,0.8,0.8666666666666667,0.8666666666666667,0.9333333333333333,1.0]","AUC":"0.5666666666666667","MacroAccuracy":"0.52", ......

4.2.4.2 windowOutput

window|{"PRC":"0.7638888888888888","SensitivityArray":"[0.3333333333333333,0.3333333333333333,0.6666666666666666,1.0,1.0,1.0]","ConfusionMatrix":"[[3,2],[0,0]]","MacroRecall":"0.5","MacroSpecificity":"0.5","FalsePositiveRateArray":"[0.0,0.5,0.5,0.5,1.0,1.0]","TruePositiveRateArray":"[0.3333333333333333,0.3333333333333333,0.6666666666666666,1.0,1.0,1.0]","AUC":"0.6666666666666666","MacroAccuracy":"0.6","RecallArray":"[0.3333333333333333,0.3333333333333333,0.6666666666666666,1.0,1.0,1.0]","KappaArray":"[0.28571428571428564,-0.15384615384615377,0.1666666666666666,0.5454545454545455,0.0,0.0]","MicroFalseNegativeRate":"0.4","WeightedRecall":"0.6","WeightedPrecision":"0.36","Recall":"1.0","MacroPrecision":"0.3",......

0xFF 參考

[[白話解析] 通過實例來梳理概念 ：準確率 (Accuracy)、精準率(Precision)、召回率(Recall) 和 F值(F-Measure)](

本站聲明:網站內容來源於博客園,如有侵權,請聯繫我們,我們將及時處理

【其他文章推薦】

※廣告預算用在刀口上，台北網頁設計公司幫您達到更多曝光效益

※新北清潔公司,居家、辦公、裝潢細清專業服務

※別再煩惱如何寫文案,掌握八大原則!

※教你寫出一流的銷售文案?

※超省錢租車方案

標籤：

您可能也會喜歡…