From 56e7a202b492fa26137b381eea73e4c773df069d Mon Sep 17 00:00:00 2001 From: Douglas Rumbaugh Date: Tue, 30 Apr 2024 14:50:53 -0400 Subject: VPTree BSM Benchmark --- benchmarks/vldb/vptree_bsm_bench.cpp | 74 ++++++++++++++++++++++++++++++++++++ 1 file changed, 74 insertions(+) create mode 100644 benchmarks/vldb/vptree_bsm_bench.cpp (limited to 'benchmarks') diff --git a/benchmarks/vldb/vptree_bsm_bench.cpp b/benchmarks/vldb/vptree_bsm_bench.cpp new file mode 100644 index 0000000..0798ec2 --- /dev/null +++ b/benchmarks/vldb/vptree_bsm_bench.cpp @@ -0,0 +1,74 @@ +/* + * + */ + +#define ENABLE_TIMER + +#include "vptree_bsm.h" +#include "file_util.h" +#include "standard_benchmarks.h" +#include "query/knn.h" + +#include + +#include "psu-util/timer.h" + + +typedef Word2VecRec Rec; + +typedef BSMVPTree Shard; +typedef de::knn::Parms QP; +typedef psudb::bsm::BentleySaxe Ext; + +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 extension = new Ext(); + gsl_rng * rng = gsl_rng_alloc(gsl_rng_mt19937); + + fprintf(stderr, "[I] Reading data file...\n"); + auto data = read_vector_file(d_fname, n); + auto queries = read_knn_queries(q_fname, 1000); + + fprintf(stderr, "[I] Warming up structure...\n"); + /* warmup structure w/ 10% of records */ + size_t warmup = .1 * n; + insert_records(extension, 0, warmup, data); + + TIMER_INIT(); + + fprintf(stderr, "[I] Running Insertion Benchmark\n"); + TIMER_START(); + insert_records(extension, warmup, data.size(), data); + 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(extension, 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 extension; + fflush(stderr); + fflush(stdout); +} + -- cgit v1.2.3