Created
October 18, 2020 05:22
-
-
Save andrwng/bae4a4696eabb501b3b69f8db263a745 to your computer and use it in GitHub Desktop.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import kudu | |
import pandas as pd | |
import pyarrow as pa | |
import time | |
def main(): | |
client = kudu.connect("0.0.0.0", 8764) | |
num_rows = 100000 | |
num_rows_per_batch = 1000 | |
num_non_pks = 100 | |
table_name = 'default.{}_non_keys'.format(num_non_pks) | |
if not client.table_exists(table_name): | |
builder = kudu.schema_builder() | |
builder.add_column('key').type(kudu.int64).nullable(False) | |
for i in range(num_non_pks): | |
builder.add_column('col{}'.format(i)).type(kudu.string) | |
builder.set_primary_keys(['key']) | |
schema = builder.build() | |
partitioning = kudu.client.Partitioning().add_hash_partitions(column_names=['key'], num_buckets=2) | |
client.create_table(table_name, schema, partitioning) | |
print("Created table {}".format(table_name)) | |
table = client.table(table_name) | |
session = client.new_session() | |
for i in range(num_rows): | |
op = table.new_insert() | |
op[0] = i | |
for i in range(num_non_pks): | |
op[1 + i] = "string-{}".format(i) | |
session.apply(op) | |
if i % num_rows_per_batch: | |
session.flush() | |
num_trials = 5 | |
tuple_total_time = 0 | |
arrow_total_time = 0 | |
table = client.table(table_name) | |
for i in range(num_trials): | |
scanner = table.scanner().open() | |
col_scanner = table.scanner().enable_columnar_layout().open() | |
start_tuples = time.perf_counter() | |
scanner.to_pandas() | |
start_arrow = time.perf_counter() | |
col_scanner.to_pandas() | |
end = time.perf_counter() | |
tuple_total_time += (start_arrow - start_tuples) | |
arrow_total_time += (end - start_arrow) | |
print("Materialized tuple DataFrame with {} rows in {} secs, averaged across {} runs".format( | |
num_rows, tuple_total_time / num_trials, num_trials)) | |
print("Materialized arrow DataFrame with {} rows in {} secs, averaged across {} runs".format( | |
num_rows, arrow_total_time / num_trials, num_trials)) | |
if __name__ == '__main__': | |
main() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment