Help Center/ Cloud Search Service/ User Guide (Paris and Amsterdam Regions)/ Vector Retrieval/ Sample Code for Vector Search on a Client
Updated on 2023-06-20 GMT+08:00
View PDF
Share

Sample Code for Vector Search on a Client

Elasticsearch provides standard REST APIs and clients developed using Java, Python, and Go.

Based on the open-source dataset SIFT1M (http://corpus-texmex.irisa.fr/) and Python Elasticsearch client, this section provides a code snippet for creating a vector index, importing vector data, and querying vector data on the client.

Prerequisites

The Python dependency package has been installed on the client. If it is not installed, run the following commands to install it:

pip install numpy
pip install elasticsearch==7.6.0

Sample Code

  1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
 
import numpy as np
import time
import json

from concurrent.futures import ThreadPoolExecutor, wait
from elasticsearch import Elasticsearch
from elasticsearch import helpers

endpoint = 'http://xxx.xxx.xxx.xxx:9200/'

# Construct an Elasticsearch client object
es = Elasticsearch(endpoint)

# Index mapping information
index_mapping = '''
{
  "settings": {
    "index": {
      "vector": "true"
    }
  },
  "mappings": {
    "properties": {
      "my_vector": {
        "type": "vector",
        "dimension": 128,
        "indexing": true,
        "algorithm": "GRAPH",
        "metric": "euclidean"
      }
    }
  }
}
'''

# Create an index.
def create_index(index_name, mapping):
    res = es.indices.create(index=index_name, ignore=400, body=mapping)
    print(res)

# Delete an index.
def delete_index(index_name):
    res = es.indices.delete(index=index_name)
    print(res)


# Refresh indexes.
def refresh_index(index_name):
    res = es.indices.refresh(index=index_name)
    print(res)


# Merge index segments.
def merge_index(index_name, seg_cnt=1):
    start = time.time()
    es.indices.forcemerge(index=index_name, max_num_segments=seg_cnt, request_timeout=36000)
    print(f" Complete the merge within {time.time() - start} seconds")


# Load vector data.
def load_vectors(file_name):
    fv = np.fromfile(file_name, dtype=np.float32)
    dim = fv.view(np.int32)[0]
    vectors = fv.reshape(-1, 1 + dim)[:, 1:]
    return vectors


# Load the ground_truth data.
def load_gts(file_name):
    fv = np.fromfile(file_name, dtype=np.int32)
    dim = fv.view(np.int32)[0]
    gts = fv.reshape(-1, 1 + dim)[:, 1:]
    return gts


def partition(ls, size):
    return [ls[i:i + size] for i in range(0, len(ls), size)]


# Write vector data.
def write_index(index_name, vec_file):
    pool = ThreadPoolExecutor(max_workers=8)
    tasks = []

    vectors = load_vectors(vec_file)
    bulk_size = 1000
    partitions = partition(vectors, bulk_size)

    start = time.time()
    start_id = 0
    for vecs in partitions:
        tasks.append(pool.submit(write_bulk, index_name, vecs, start_id))
        start_id += len(vecs)
    wait(tasks)
    print(f" Complete the writing within {time.time() - start} seconds")


def write_bulk(index_name, vecs, start_id):
    actions = [
        {
            "_index": index_name,
            "my_vector": vecs[j].tolist(),
            "_id": str(j + start_id)
        }
        for j in range(len(vecs))
    ]
    helpers.bulk(es, actions, request_timeout=3600)


# Query an index.
def search_index(index_name, query_file, gt_file, k):
    print("Start query! Index name: " + index_name)

    queries = load_vectors(query_file)
    gt = load_gts(gt_file)

    took = 0
    precision = []
    for idx, query in enumerate(queries):
        hits = set()
        query_json = {
                  "size": k,
                  "_source": False,
                  "query": {
                    "vector": {
                      "my_vector": {
                        "vector": query.tolist(),
                        "topk": k
                      }
                    }
                  }
                }
        res = es.search(index=index_name, body=json.dumps(query_json))

        for hit in res['hits']['hits']:
            hits.add(int(hit['_id']))
        precision.append(len(hits.intersection(set(gt[idx, :k]))) / k)
        took += res['took']

    print("precision: " + str(sum(precision) / len(precision)))
    print(f" Complete the retrieval within {took / 1000:.2f} seconds; average took size is {took / len(queries):.2f} ms")


if __name__ == "__main__":
    vec_file = r"./data/sift/sift_base.fvecs"
    qry_file = r"./data/sift/sift_query.fvecs"
    gt_file = r"./data/sift/sift_groundtruth.ivecs"

    index = "test"
    create_index(index, index_mapping)
    write_index(index, vec_file)
    merge_index(index)
    refresh_index(index)

    search_index(index, qry_file, gt_file, 10)
Parent topic: Vector Retrieval