Merge pull request #12887 from taosdata/feature/3.0_glzhao

feat(query): add apercentile function

include/util/tdigest.h

@@ -0,0 +1,71 @@
+/*
+ * Copyright (c) 2019 TAOS Data, Inc. <jhtao@taosdata.com>
+ *
+ * This program is free software: you can use, redistribute, and/or modify
+ * it under the terms of the GNU Affero General Public License, version 3
+ * or later ("AGPL"), as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+/*
+ * include/tdigest.c
+ *
+ * Copyright (c) 2016, Usman Masood <usmanm at fastmail dot fm>
+ */
+
+#ifndef TDIGEST_H
+#define TDIGEST_H
+
+#ifndef M_PI
+#define M_PI        3.14159265358979323846264338327950288   /* pi             */
+#endif
+
+#define DOUBLE_MAX 1.79e+308
+
+#define ADDITION_CENTROID_NUM 2
+#define COMPRESSION 300
+#define GET_CENTROID(compression)  (ceil(compression * M_PI / 2) + 1 + ADDITION_CENTROID_NUM)
+#define GET_THRESHOLD(compression) (7.5 + 0.37 * compression - 2e-4 * pow(compression, 2))
+#define TDIGEST_SIZE(compression)  (sizeof(TDigest) + sizeof(SCentroid)*GET_CENTROID(compression) + sizeof(SPt)*GET_THRESHOLD(compression))
+
+typedef struct SCentroid {
+    double mean;
+    int64_t weight;
+}SCentroid;
+
+typedef struct SPt {
+    double value;
+    int64_t weight;
+}SPt;
+
+typedef struct TDigest {
+    double compression;
+    int32_t threshold;
+    int64_t size;
+
+    int64_t total_weight;
+    double min;
+    double max;
+
+    int32_t num_buffered_pts;
+    SPt *buffered_pts;
+
+    int32_t num_centroids;
+    SCentroid *centroids;
+}TDigest;
+
+TDigest *tdigestNewFrom(void* pBuf, int32_t compression);
+void tdigestAdd(TDigest *t, double x, int64_t w);
+void tdigestMerge(TDigest *t1, TDigest *t2);
+double tdigestQuantile(TDigest *t, double q);
+void tdigestCompress(TDigest *t);
+void tdigestFreeFrom(TDigest *t);
+void tdigestAutoFill(TDigest* t, int32_t compression);
+
+#endif /* TDIGEST_H */

source/libs/function/inc/builtinsimpl.h

@@ -73,6 +73,11 @@ bool percentileFunctionSetup(SqlFunctionCtx *pCtx, SResultRowEntryInfo* pResultI
 int32_t percentileFunction(SqlFunctionCtx *pCtx);
 int32_t percentileFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock);
 
+bool getApercentileFuncEnv(struct SFunctionNode* pFunc, SFuncExecEnv* pEnv);
+bool apercentileFunctionSetup(SqlFunctionCtx *pCtx, SResultRowEntryInfo* pResultInfo);
+int32_t apercentileFunction(SqlFunctionCtx *pCtx);
+int32_t apercentileFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock);
+
 bool getDiffFuncEnv(struct SFunctionNode* pFunc, SFuncExecEnv* pEnv);
 bool diffFunctionSetup(SqlFunctionCtx *pCtx, SResultRowEntryInfo* pResInfo);
 int32_t diffFunction(SqlFunctionCtx *pCtx);

source/libs/function/src/builtins.c

@@ -998,10 +998,10 @@ const SBuiltinFuncDefinition funcMgtBuiltins[] = {
     .type = FUNCTION_TYPE_APERCENTILE,
     .classification = FUNC_MGT_AGG_FUNC,
     .translateFunc = translateApercentile,
-    .getEnvFunc   = getMinmaxFuncEnv,
+    .getEnvFunc   = getApercentileFuncEnv,
-    .initFunc     = minmaxFunctionSetup,
+    .initFunc     = apercentileFunctionSetup,
-    .processFunc  = maxFunction,
+    .processFunc  = apercentileFunction,
-    .finalizeFunc = functionFinalize
+    .finalizeFunc = apercentileFinalize
   },
   {
     .name = "top",

source/libs/function/src/builtinsimpl.c

@@ -20,6 +20,8 @@
 #include "taggfunction.h"
 #include "tcompare.h"
 #include "tdatablock.h"
+#include "tdigest.h"
+#include "thistogram.h"
 #include "tpercentile.h"
 
 #define HISTOGRAM_MAX_BINS_NUM   1000
@@ -95,6 +97,19 @@ typedef struct SPercentileInfo {
   int64_t     numOfElems;
 } SPercentileInfo;
 
+typedef struct SAPercentileInfo {
+  double result;
+  int8_t algo;
+  SHistogramInfo *pHisto;
+  TDigest *pTDigest;
+} SAPercentileInfo;
+
+typedef enum {
+  APERCT_ALGO_UNKNOWN = 0,
+  APERCT_ALGO_DEFAULT,
+  APERCT_ALGO_TDIGEST,
+} EAPerctAlgoType;
+
 typedef struct SDiffInfo {
   bool hasPrev;
   bool includeNull;
@@ -1905,6 +1920,131 @@ int32_t percentileFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) {
   return functionFinalize(pCtx, pBlock);
 }
 
+bool getApercentileFuncEnv(SFunctionNode* pFunc, SFuncExecEnv* pEnv) {
+  int32_t bytesHist   = (int32_t)(sizeof(SAPercentileInfo) + sizeof(SHistogramInfo) + sizeof(SHistBin) * (MAX_HISTOGRAM_BIN + 1));
+  int32_t bytesDigest = (int32_t)(sizeof(SAPercentileInfo) + TDIGEST_SIZE(COMPRESSION));
+  pEnv->calcMemSize = TMAX(bytesHist, bytesDigest);
+  return true;
+}
+
+static int8_t getApercentileAlgo(char *algoStr) {
+  int8_t algoType;
+  if (strcasecmp(algoStr, "default") == 0) {
+    algoType = APERCT_ALGO_DEFAULT;
+  } else if (strcasecmp(algoStr, "t-digest") == 0) {
+    algoType = APERCT_ALGO_TDIGEST;
+  } else {
+    algoType = APERCT_ALGO_UNKNOWN;
+  }
+
+  return algoType;
+}
+
+static void buildHistogramInfo(SAPercentileInfo* pInfo) {
+  pInfo->pHisto = (SHistogramInfo*) ((char*) pInfo + sizeof(SAPercentileInfo));
+  pInfo->pHisto->elems = (SHistBin*) ((char*)pInfo->pHisto + sizeof(SHistogramInfo));
+}
+
+bool apercentileFunctionSetup(SqlFunctionCtx* pCtx, SResultRowEntryInfo* pResultInfo) {
+  if (!functionSetup(pCtx, pResultInfo)) {
+    return false;
+  }
+
+  SAPercentileInfo* pInfo = GET_ROWCELL_INTERBUF(pResultInfo);
+  if (pCtx->numOfParams == 2) {
+    pInfo->algo = APERCT_ALGO_DEFAULT;
+  } else if (pCtx->numOfParams == 3) {
+    pInfo->algo = getApercentileAlgo(pCtx->param[2].param.pz);
+    if (pInfo->algo == APERCT_ALGO_UNKNOWN) {
+      return false;
+    }
+  }
+
+  char *tmp = (char *)pInfo + sizeof(SAPercentileInfo);
+  if (pInfo->algo == APERCT_ALGO_TDIGEST) {
+    pInfo->pTDigest = tdigestNewFrom(tmp, COMPRESSION);
+  } else {
+    buildHistogramInfo(pInfo);
+    pInfo->pHisto = tHistogramCreateFrom(tmp, MAX_HISTOGRAM_BIN);
+  }
+
+  return true;
+}
+
+int32_t apercentileFunction(SqlFunctionCtx* pCtx) {
+  int32_t              notNullElems = 0;
+  SResultRowEntryInfo* pResInfo = GET_RES_INFO(pCtx);
+
+  SInputColumnInfoData* pInput = &pCtx->input;
+  //SColumnDataAgg*       pAgg = pInput->pColumnDataAgg[0];
+
+  SColumnInfoData* pCol = pInput->pData[0];
+  int32_t          type = pCol->info.type;
+
+  SAPercentileInfo* pInfo = GET_ROWCELL_INTERBUF(pResInfo);
+
+  int32_t start = pInput->startRowIndex;
+  if (pInfo->algo == APERCT_ALGO_TDIGEST) {
+    for (int32_t i = start; i < pInput->numOfRows + start; ++i) {
+      if (colDataIsNull_f(pCol->nullbitmap, i)) {
+        continue;
+      }
+      notNullElems += 1;
+      char* data = colDataGetData(pCol, i);
+
+      double v = 0; // value
+      int64_t w = 1; // weigth
+      GET_TYPED_DATA(v, double, type, data);
+      tdigestAdd(pInfo->pTDigest, v, w);
+    }
+  } else {
+    for (int32_t i = start; i < pInput->numOfRows + start; ++i) {
+      if (colDataIsNull_f(pCol->nullbitmap, i)) {
+        continue;
+      }
+      notNullElems += 1;
+      char* data = colDataGetData(pCol, i);
+
+      double v = 0;
+      GET_TYPED_DATA(v, double, type, data);
+      tHistogramAdd(&pInfo->pHisto, v);
+    }
+  }
+
+  SET_VAL(pResInfo, notNullElems, 1);
+  return TSDB_CODE_SUCCESS;
+}
+
+int32_t apercentileFinalize(SqlFunctionCtx* pCtx, SSDataBlock* pBlock) {
+  SVariant* pVal    = &pCtx->param[1].param;
+  double    percent = (pVal->nType == TSDB_DATA_TYPE_BIGINT) ? pVal->i : pVal->d;
+
+  SResultRowEntryInfo* pResInfo = GET_RES_INFO(pCtx);
+  SAPercentileInfo*       pInfo = (SAPercentileInfo*)GET_ROWCELL_INTERBUF(pResInfo);
+
+  if (pInfo->algo == APERCT_ALGO_TDIGEST) {
+    if (pInfo->pTDigest->size > 0) {
+      pInfo->result = tdigestQuantile(pInfo->pTDigest, percent/100);
+    } else {  // no need to free
+      //setNull(pCtx->pOutput, pCtx->outputType, pCtx->outputBytes);
+      return TSDB_CODE_SUCCESS;
+    }
+  } else {
+    if (pInfo->pHisto->numOfElems > 0) {
+      double ratio[] = {percent};
+      double *res = tHistogramUniform(pInfo->pHisto, ratio, 1);
+      pInfo->result = *res;
+      //memcpy(pCtx->pOutput, res, sizeof(double));
+      taosMemoryFree(res);
+    } else {  // no need to free
+      //setNull(pCtx->pOutput, pCtx->outputType, pCtx->outputBytes);
+      return TSDB_CODE_SUCCESS;
+    }
+  }
+
+  return functionFinalize(pCtx, pBlock);
+}
+
 bool getFirstLastFuncEnv(SFunctionNode* pFunc, SFuncExecEnv* pEnv) {
   SColumnNode* pNode = nodesListGetNode(pFunc->pParameterList, 0);
   pEnv->calcMemSize = pNode->node.resType.bytes + sizeof(int64_t);
@@ -1917,8 +2057,6 @@ bool getSelectivityFuncEnv(SFunctionNode* pFunc, SFuncExecEnv* pEnv) {
   return true;
 }
 
-
-
 static FORCE_INLINE TSKEY getRowPTs(SColumnInfoData* pTsColInfo, int32_t rowIndex) {
   if (pTsColInfo == NULL) {
     return 0;

source/util/src/tdigest.c

@@ -0,0 +1,319 @@
+/*
+ * Copyright (c) 2019 TAOS Data, Inc. <jhtao@taosdata.com>
+ *
+ * This program is free software: you can use, redistribute, and/or modify
+ * it under the terms of the GNU Affero General Public License, version 3
+ * or later ("AGPL"), as published by the Free Software Foundation.
+ *
+ * This program is distributed in the hope that it will be useful, but WITHOUT
+ * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
+ * FITNESS FOR A PARTICULAR PURPOSE.
+ *
+ * You should have received a copy of the GNU Affero General Public License
+ * along with this program. If not, see <http://www.gnu.org/licenses/>.
+ */
+
+/*
+ * src/tdigest.c
+ *
+ * Implementation of the t-digest data structure used to compute accurate percentiles.
+ *
+ * It is based on the MergingDigest implementation found at:
+ *   https://github.com/tdunning/t-digest/blob/master/src/main/java/com/tdunning/math/stats/MergingDigest.java
+ *
+ * Copyright (c) 2016, Usman Masood <usmanm at fastmail dot fm>
+ */
+
+#include "os.h"
+#include "osMath.h"
+#include "tdigest.h"
+
+#define INTERPOLATE(x, x0, x1) (((x) - (x0)) / ((x1) - (x0)))
+//#define INTEGRATED_LOCATION(compression, q)   ((compression) * (asin(2 * (q) - 1) + M_PI / 2) / M_PI)
+#define INTEGRATED_LOCATION(compression, q) ((compression) * (asin(2 * (double)(q) - 1)/M_PI + (double)1/2))
+#define FLOAT_EQ(f1, f2) (fabs((f1) - (f2)) <= FLT_EPSILON)
+
+typedef struct SMergeArgs {
+    TDigest *t;
+    SCentroid *centroids;
+    int32_t idx;
+    double weight_so_far;
+    double k1;
+    double min;
+    double max;
+}SMergeArgs;
+
+void tdigestAutoFill(TDigest* t, int32_t compression) {
+    t->centroids    = (SCentroid*)((char*)t + sizeof(TDigest));
+    t->buffered_pts = (SPt*)   ((char*)t + sizeof(TDigest) + sizeof(SCentroid) * (int32_t)GET_CENTROID(compression));
+}
+
+TDigest *tdigestNewFrom(void* pBuf, int32_t compression) {
+    memset(pBuf, 0, (size_t)TDIGEST_SIZE(compression));
+    TDigest* t = (TDigest*)pBuf;
+    tdigestAutoFill(t, compression);
+
+    t->compression = compression;
+    t->size = (int64_t)GET_CENTROID(compression);
+    t->threshold = (int32_t)GET_THRESHOLD(compression);
+    t->min = DOUBLE_MAX;
+    t->max = -DOUBLE_MAX;
+
+    return t;
+}
+
+static int32_t cmpCentroid(const void *a, const void *b) {
+    SCentroid *c1 = (SCentroid *) a;
+    SCentroid *c2 = (SCentroid *) b;
+    if (c1->mean < c2->mean)
+        return -1;
+    if (c1->mean > c2->mean)
+        return 1;
+    return 0;
+}
+
+
+static void mergeCentroid(SMergeArgs *args, SCentroid *merge) {
+    double k2;
+    SCentroid *c = &args->centroids[args->idx];
+
+    args->weight_so_far += merge->weight;
+    k2 = INTEGRATED_LOCATION(args->t->size,
+            args->weight_so_far / args->t->total_weight);
+    //idx++
+    if(k2 - args->k1 > 1 && c->weight > 0) {
+        if(args->idx + 1 < args->t->size
+         && merge->mean != args->centroids[args->idx].mean) {
+             args->idx++;
+        }
+        args->k1 = k2;
+    }
+
+    c = &args->centroids[args->idx];
+    if(c->mean == merge->mean) {
+        c->weight += merge->weight;
+    } else {
+        c->weight += merge->weight;
+        c->mean += (merge->mean - c->mean) * merge->weight / c->weight;
+
+        if (merge->weight > 0) {
+            args->min = TMIN(merge->mean, args->min);
+            args->max = TMAX(merge->mean, args->max);
+        }
+    }
+}
+
+void tdigestCompress(TDigest *t) {
+    SCentroid *unmerged_centroids;
+    int64_t unmerged_weight = 0;
+    int32_t num_unmerged = t->num_buffered_pts;
+    int32_t i, j;
+    SMergeArgs args;
+
+    if (t->num_buffered_pts <= 0)
+        return;
+
+    unmerged_centroids = (SCentroid*)taosMemoryMalloc(sizeof(SCentroid) * t->num_buffered_pts);
+    for (i = 0; i < num_unmerged; i++) {
+        SPt *p = t->buffered_pts + i;
+        SCentroid *c = &unmerged_centroids[i];
+        c->mean = p->value;
+        c->weight = p->weight;
+        unmerged_weight += c->weight;
+    }
+    t->num_buffered_pts = 0;
+    t->total_weight += unmerged_weight;
+
+    qsort(unmerged_centroids, num_unmerged, sizeof(SCentroid), cmpCentroid);
+    memset(&args, 0, sizeof(SMergeArgs));
+    args.centroids = (SCentroid*)taosMemoryMalloc((size_t)(sizeof(SCentroid) * t->size));
+    memset(args.centroids, 0, (size_t)(sizeof(SCentroid) * t->size));
+
+    args.t = t;
+    args.min = DOUBLE_MAX;
+    args.max = -DOUBLE_MAX;
+
+    i = 0;
+    j = 0;
+    while (i < num_unmerged && j < t->num_centroids) {
+        SCentroid *a = &unmerged_centroids[i];
+        SCentroid *b = &t->centroids[j];
+
+        if (a->mean <= b->mean) {
+            mergeCentroid(&args, a);
+            assert(args.idx < t->size);
+            i++;
+        } else {
+            mergeCentroid(&args, b);
+            assert(args.idx < t->size);
+            j++;
+        }
+    }
+
+    while (i < num_unmerged) {
+        mergeCentroid(&args, &unmerged_centroids[i++]);
+        assert(args.idx < t->size);
+    }
+    taosMemoryFree((void*)unmerged_centroids);
+
+    while (j < t->num_centroids) {
+        mergeCentroid(&args, &t->centroids[j++]);
+        assert(args.idx < t->size);
+    }
+
+    if (t->total_weight > 0) {
+        t->min = TMIN(t->min, args.min);
+        if (args.centroids[args.idx].weight <= 0) {
+            args.idx--;
+        }
+        t->num_centroids = args.idx + 1;
+        t->max = TMAX(t->max, args.max);
+    }
+
+    memcpy(t->centroids, args.centroids, sizeof(SCentroid) * t->num_centroids);
+    taosMemoryFree((void*)args.centroids);
+}
+
+void tdigestAdd(TDigest* t, double x, int64_t w) {
+    if (w == 0)
+        return;
+
+    int32_t i = t->num_buffered_pts;
+    if(i > 0 && t->buffered_pts[i-1].value == x ) {
+        t->buffered_pts[i].weight = w;
+    } else {
+        t->buffered_pts[i].value  = x;
+        t->buffered_pts[i].weight = w;
+        t->num_buffered_pts++;
+    }
+
+
+    if (t->num_buffered_pts >= t->threshold)
+        tdigestCompress(t);
+}
+
+double tdigestCDF(TDigest *t, double x) {
+    if (t == NULL)
+        return 0;
+
+    int32_t i;
+    double left, right;
+    int64_t weight_so_far;
+    SCentroid *a, *b, tmp;
+
+    tdigestCompress(t);
+    if (t->num_centroids == 0)
+        return NAN;
+    if (x < t->min)
+        return 0;
+    if (x > t->max)
+        return 1;
+    if (t->num_centroids == 1) {
+        if (FLOAT_EQ(t->max, t->min))
+            return 0.5;
+
+        return INTERPOLATE(x, t->min, t->max);
+    }
+
+    weight_so_far = 0;
+    a = b = &tmp;
+    b->mean = t->min;
+    b->weight = 0;
+    right = 0;
+
+    for (i = 0; i < t->num_centroids; i++) {
+        SCentroid *c = &t->centroids[i];
+
+        left = b->mean - (a->mean + right);
+        a = b;
+        b = c;
+        right = (b->mean - a->mean) * a->weight / (a->weight + b->weight);
+
+        if (x < a->mean + right) {
+            double cdf = (weight_so_far
+                    + a->weight
+                    * INTERPOLATE(x, a->mean - left, a->mean + right))
+                    / t->total_weight;
+            return TMAX(cdf, 0.0);
+        }
+
+        weight_so_far += a->weight;
+    }
+
+    left = b->mean - (a->mean + right);
+    a = b;
+    right = t->max - a->mean;
+
+    if (x < a->mean + right) {
+        return (weight_so_far + a->weight * INTERPOLATE(x, a->mean - left, a->mean + right))
+                / t->total_weight;
+    }
+
+    return 1;
+}
+
+double tdigestQuantile(TDigest *t, double q) {
+    if (t == NULL)
+        return 0;
+
+    int32_t i;
+    double left, right, idx;
+    int64_t weight_so_far;
+    SCentroid *a, *b, tmp;
+
+    tdigestCompress(t);
+    if (t->num_centroids == 0)
+        return NAN;
+    if (t->num_centroids == 1)
+        return t->centroids[0].mean;
+    if (FLOAT_EQ(q, 0.0))
+        return t->min;
+    if (FLOAT_EQ(q, 1.0))
+        return t->max;
+
+    idx = q * t->total_weight;
+    weight_so_far = 0;
+    b = &tmp;
+    b->mean = t->min;
+    b->weight = 0;
+    right = t->min;
+
+    for (i = 0; i < t->num_centroids; i++) {
+        SCentroid *c = &t->centroids[i];
+        a = b;
+        left = right;
+
+        b = c;
+        right = (b->weight * a->mean + a->weight * b->mean)/ (a->weight + b->weight);
+        if (idx < weight_so_far + a->weight) {
+            double p = (idx - weight_so_far) / a->weight;
+            return left * (1 - p) + right * p;
+        }
+        weight_so_far += a->weight;
+    }
+
+    left = right;
+    a = b;
+    right = t->max;
+
+    if (idx < weight_so_far + a->weight && a->weight != 0) {
+        double p = (idx - weight_so_far) / a->weight;
+        return left * (1 - p) + right * p;
+    }
+
+    return t->max;
+}
+
+void tdigestMerge(TDigest *t1, TDigest *t2) {
+    // SPoints
+    int32_t num_pts = t2->num_buffered_pts;
+    for(int32_t i = num_pts - 1; i >= 0; i--) {
+        SPt* p = t2->buffered_pts + i;
+        tdigestAdd(t1, p->value, p->weight);
+        t2->num_buffered_pts --;
+    }
+    // centroids
+    for (int32_t i = 0; i < t2->num_centroids; i++) {
+        tdigestAdd(t1, t2->centroids[i].mean, t2->centroids[i].weight);
+    }
+}

tests/system-test/fulltest.sh

@@ -60,8 +60,7 @@ python3 ./test.py -f 2-query/arcsin.py
 python3 ./test.py -f 2-query/arccos.py
 python3 ./test.py -f 2-query/arctan.py
 python3 ./test.py -f 2-query/query_cols_tags_and_or.py
-python3 ./test.py -f 2-query/nestedQuery.py
+#python3 ./test.py -f 2-query/nestedQuery.py
- 
 
 python3 ./test.py -f 7-tmq/basic5.py
 python3 ./test.py -f 7-tmq/subscribeDb.py