More than 5 years have passed since last update.
Python 3.7がリリースされていたので、Benchmarkerのサンプルコードで簡単にベンチマークした結果です。
ベンチマーク結果
3.6
$ python3 example.py -o result3_6.json
## benchmarker: release 4.0.1 (for python)
## python version: 3.6.5
## python compiler: GCC 4.2.1 Compatible Apple LLVM 9.0.0 (clang-900.0.39.2)
## python platform: Darwin-16.7.0-x86_64-i386-64bit
## python executable: /usr/local/opt/python/bin/python3.6
## cpu model: Intel(R) Core(TM) i5-7360U CPU @ 2.30GHz
## parameters: loop=1000000, cycle=1, extra=0
## real (total = user + sys)
(Empty) 0.0277 0.0300 0.0300 0.0000
join 0.2340 0.2300 0.2300 0.0000
concat 0.2714 0.2700 0.2700 0.0000
format 0.3307 0.3200 0.3200 0.0000
## Ranking real
join 0.2340 (100.0) ********************
concat 0.2714 ( 86.2) *****************
format 0.3307 ( 70.8) **************
## Matrix real [01] [02] [03]
[01] join 0.2340 100.0 116.0 141.3
[02] concat 0.2714 86.2 100.0 121.9
[03] format 0.3307 70.8 82.0 100.0
3.7
$ python3 example.py -o result3_7.json
## benchmarker: release 4.0.1 (for python)
## python version: 3.7.0
## python compiler: Clang 9.0.0 (clang-900.0.39.2)
## python platform: Darwin-16.7.0-x86_64-i386-64bit
## python executable: /usr/local/opt/python/bin/python3.7
## cpu model: Intel(R) Core(TM) i5-7360U CPU @ 2.30GHz
## parameters: loop=1000000, cycle=1, extra=0
## real (total = user + sys)
(Empty) 0.0258 0.0200 0.0200 0.0000
join 0.2265 0.2400 0.2300 0.0100
concat 0.3118 0.3200 0.3200 0.0000
format 0.3388 0.3400 0.3400 0.0000
## Ranking real
join 0.2265 (100.0) ********************
concat 0.3118 ( 72.7) ***************
format 0.3388 ( 66.9) *************
## Matrix real [01] [02] [03]
[01] join 0.2265 100.0 137.6 149.6
[02] concat 0.3118 72.7 100.0 108.7
[03] format 0.3388 66.9 92.0 100.0
所感
文字列結合がそこそこ遅くなっているのが目立つ。
3.8も出たので計測
計測マシンは変わってしまったので、参考値としてですが同様に計測しました。
$ python3.8 example.py -o result3_8.json
## benchmarker: release 4.0.1 (for python)
## python version: 3.8.4
## python compiler: Clang 11.0.3 (clang-1103.0.32.62)
## python platform: macOS-10.15.5-x86_64-i386-64bit
## python executable: /usr/local/opt/python@3.8/bin/python3.8
## cpu model: Intel(R) Core(TM) i5-8279U CPU @ 2.40GHz
## parameters: loop=1000000, cycle=1, extra=0
## real (total = user + sys)
(Empty) 0.0250 0.0200 0.0200 0.0000
join 0.2017 0.2100 0.2100 0.0000
concat 0.2530 0.2600 0.2600 0.0000
format 0.3028 0.3000 0.3000 0.0000
## Ranking real
join 0.2017 (100.0) ********************
concat 0.2530 ( 79.7) ****************
format 0.3028 ( 66.6) *************
## Matrix real [01] [02] [03]
[01] join 0.2017 100.0 125.5 150.1
[02] concat 0.2530 79.7 100.0 119.7
[03] format 0.3028 66.6 83.6 100.0
3.9も出たので計測
3.8と同じマシンで計測しました。
$ python3.9 example.py -o result3_9.json
## benchmarker: release 4.0.1 (for python)
## python version: 3.9.0
## python compiler: Clang 11.0.3 (clang-1103.0.32.62)
## python platform: macOS-10.15.7-x86_64-i386-64bit
## python executable: /usr/local/opt/python@3.9/bin/python3.9
## cpu model: Intel(R) Core(TM) i5-8279U CPU @ 2.40GHz
## parameters: loop=1000000, cycle=1, extra=0
## real (total = user + sys)
(Empty) 0.0243 0.0200 0.0200 0.0000
join 0.2200 0.2300 0.2300 0.0000
concat 0.2766 0.2700 0.2700 0.0000
format 0.3161 0.3200 0.3200 0.0000
## Ranking real
join 0.2200 (100.0) ********************
concat 0.2766 ( 79.5) ****************
format 0.3161 ( 69.6) **************
## Matrix real [01] [02] [03]
[01] join 0.2200 100.0 125.7 143.7
[02] concat 0.2766 79.5 100.0 114.3
[03] format 0.3161 69.6 87.5 100.0
3.8よりは遅くなっている感じ。
Register as a new user and use Qiita more conveniently
- You get articles that match your needs
- You can efficiently read back useful information
- You can use dark theme