From 4a1dde3148e0e84b47c884bc0bb69c60678b4558 Mon Sep 17 00:00:00 2001
From: Douglas Rumbaugh <dbr4@psu.edu>
Date: Mon, 22 Apr 2024 15:09:07 -0400
Subject: Benchmark update+reorganization

The Alex benchmark isn't updated yet.
---
 benchmarks/vldb/mtree_bench.cpp | 80 +++++++++++++++++++++++++++++++++++++++++
 1 file changed, 80 insertions(+)
 create mode 100644 benchmarks/vldb/mtree_bench.cpp

(limited to 'benchmarks/vldb/mtree_bench.cpp')

diff --git a/benchmarks/vldb/mtree_bench.cpp b/benchmarks/vldb/mtree_bench.cpp
new file mode 100644
index 0000000..35f56be
--- /dev/null
+++ b/benchmarks/vldb/mtree_bench.cpp
@@ -0,0 +1,80 @@
+/*
+ *
+ */
+
+#define ENABLE_TIMER
+
+#include "query/knn.h"
+#include "file_util.h"
+#include "standard_benchmarks.h"
+
+#include <gsl/gsl_rng.h>
+
+#include "psu-util/timer.h"
+
+
+typedef Word2VecRec Rec;
+typedef de::knn::Parms<Rec> QP;
+
+void usage(char *progname) {
+    fprintf(stderr, "%s reccnt datafile queryfile\n", progname);
+}
+
+int main(int argc, char **argv) {
+
+    if (argc < 4) {
+        usage(argv[0]);
+        exit(EXIT_FAILURE);
+    }
+
+    size_t n = atol(argv[1]);
+    std::string d_fname = std::string(argv[2]);
+    std::string q_fname = std::string(argv[3]);
+
+    auto mtree = new MTree();
+    gsl_rng * rng = gsl_rng_alloc(gsl_rng_mt19937);
+    
+    fprintf(stderr, "[I] Reading data file...\n");
+    auto data = read_vector_file<Rec, 300>(d_fname, n);
+
+    fprintf(stderr, "[I] Generating delete vector\n");
+    std::vector<size_t> to_delete(n * delete_proportion);
+    size_t j=0;
+    for (size_t i=0; i<data.size() && j<to_delete.size(); i++) {
+        if (gsl_rng_uniform(rng) <= delete_proportion) {
+            to_delete[j++] = i;
+        } 
+    }
+    fprintf(stderr, "[I] Reading Queries\n");
+    auto queries = read_knn_queries<QP>(q_fname, 10);
+
+    fprintf(stderr, "[I] Warming up structure...\n");
+    /* warmup structure w/ 10% of records */
+    size_t warmup = .1 * n;
+    size_t delete_idx = 0;
+    insert_records<MTree, Rec>(mtree, 0, warmup, data, to_delete, delete_idx, false, rng);
+
+    TIMER_INIT();
+
+    fprintf(stderr, "[I] Running Insertion Benchmark\n");
+    TIMER_START();
+    insert_records<MTree, Rec>(mtree, warmup, data.size(), data, to_delete, delete_idx, true, rng);
+    TIMER_STOP();
+
+    auto insert_latency = TIMER_RESULT();
+    size_t insert_throughput = (size_t) ((double) (n - warmup) / (double) insert_latency * 1e9);
+
+    fprintf(stderr, "[I] Running Query Benchmark\n");
+    TIMER_START();
+    run_queries<MTree, QP>(mtree, queries);
+    TIMER_STOP();
+
+    auto query_latency = TIMER_RESULT() / queries.size();
+
+    fprintf(stdout, "%ld\t%ld\n", insert_throughput, query_latency);
+
+    gsl_rng_free(rng);
+    delete mtree;
+    fflush(stderr);
+}
+
-- 
cgit v1.2.3


From 47a386b50d904d3f1b7ce3cfc13c29ea96dd1e43 Mon Sep 17 00:00:00 2001
From: Douglas Rumbaugh <dbr4@psu.edu>
Date: Tue, 30 Apr 2024 14:18:08 -0400
Subject: Added VPTree BSM benchmark

---
 benchmarks/vldb/mtree_bench.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'benchmarks/vldb/mtree_bench.cpp')

diff --git a/benchmarks/vldb/mtree_bench.cpp b/benchmarks/vldb/mtree_bench.cpp
index 35f56be..60425da 100644
--- a/benchmarks/vldb/mtree_bench.cpp
+++ b/benchmarks/vldb/mtree_bench.cpp
@@ -46,7 +46,7 @@ int main(int argc, char **argv) {
         } 
     }
     fprintf(stderr, "[I] Reading Queries\n");
-    auto queries = read_knn_queries<QP>(q_fname, 10);
+    auto queries = read_knn_queries<QP>(q_fname, 1000);
 
     fprintf(stderr, "[I] Warming up structure...\n");
     /* warmup structure w/ 10% of records */
-- 
cgit v1.2.3


From ab0ab297959fcca370e80670e17f90a780607a80 Mon Sep 17 00:00:00 2001
From: Douglas Rumbaugh <dbr4@psu.edu>
Date: Fri, 10 May 2024 18:35:30 -0400
Subject: MTree structure size

---
 benchmarks/vldb/mtree_bench.cpp | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

(limited to 'benchmarks/vldb/mtree_bench.cpp')

diff --git a/benchmarks/vldb/mtree_bench.cpp b/benchmarks/vldb/mtree_bench.cpp
index 60425da..cc2f41f 100644
--- a/benchmarks/vldb/mtree_bench.cpp
+++ b/benchmarks/vldb/mtree_bench.cpp
@@ -71,7 +71,9 @@ int main(int argc, char **argv) {
 
     auto query_latency = TIMER_RESULT() / queries.size();
 
-    fprintf(stdout, "%ld\t%ld\n", insert_throughput, query_latency);
+    auto size = mtree->size() - sizeof(Rec)*(data.size() - to_delete.size());
+
+    fprintf(stdout, "%ld\t%ld\t%ld\n", insert_throughput, query_latency, size);
 
     gsl_rng_free(rng);
     delete mtree;
-- 
cgit v1.2.3