Python vs Rust for Neural Networks | Nathan Goldbaum
In a previous post I introduced the MNIST dataset and the problem of classifying handwritten digits. In this post I’ll be using the code I wrote in that post to port a simple neural network implementation to rust. My goal is to explore performance and ergonomics for data science workflows in rust. The Python Implementation Chapter 1 of the book describes a very simple single-layer Neural Network t
Pythonユーザーãªã‚‰çŸ¥ã£ã¦ãŠããŸã„ã®ãƒªã‚¹ãƒˆã®ä»•çµ„㿠– Yasufumi Taniguchi – Medium
Pythonã§ãƒ—ãƒã‚°ãƒ©ãƒ を書ãã¨ãã€ã»ã¼å¿…é ˆã¨ãªã‚‹ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã§ã‚るリスト (list) ã®ä»•çµ„ã¿ã‚’紹介ã—ã¾ã™ã€‚僕自身Pythonをよã使ã†ã®ã§ã™ãŒã€ã“ã‚Œã¾ã§å®Ÿè£…ã‚’æ„è˜ã—ã¦ã“ãªã‹ã£ãŸã®ã§ã€å†…部ã®ä»•çµ„ã¿ã«ã¤ã„ã¦ã¾ã¨ã‚ã¦ã¿ã¾ã—ãŸã€‚Pythonã®ãƒªã‚¹ãƒˆã¯è¦ç´ ã®è¿½åŠ (list.append) /削除 (list.pop) ã«ã‚ˆã‚Šã€ã‚µã‚¤ã‚ºãŒå‹•çš„ã«å¤‰æ›´ã•ã‚Œã¾ã™ãŒã€ã“れらã¯O(1)ã§é«˜é€Ÿã«è¡Œã† (list.popã¯æœ«å°¾è¦ç´ ã®ã¿) ã“ã¨ãŒã§ãã¾ã™ã€‚ã“ã®è¨˜äº‹ã§ã¯ãã®ç†ç”±ã«ã¤ã„ã¦è§£èª¬ã—ã¦ã„ãã¾ã™ã€‚ Pythonã®ãƒªã‚¹ãƒˆã®ã‚ˆã†ãªç‰¹å¾´ã‚’æŒã£ãŸãƒ‡ãƒ¼ã‚¿æ§‹é€ ã¯å‹•çš„é…列 (dynamic array) ã¨å‘¼ã°ã‚Œã¾ã™ã€‚通常ã®é…列ã¯é™çš„é…列 (static array)ã¨ã—ã¦åŒºåˆ¥ã•ã‚Œã¾ã™ã€‚ ã“ã®è¨˜äº‹ã§ã¯ã€å‹•çš„é…列ã®æ¦‚è¦ã€è¦ç´ ã®è¿½åŠ /削除ã«ã¤ã„ã¦ã€MITã®è¬›ç¾© (Table Doubling, Karp-Rabin) ã‚’å…ƒã«ä»¥
Python: scikit-learn ã®ãƒã‚¸ã‚¹ãƒ†ã‚£ãƒƒã‚¯å›žå¸°ã‚’使ã£ã¦ã¿ã‚‹ - CUBE SUGAR CONTAINER
最近ã€æ„外ã¨ãƒã‚¸ã‚¹ãƒ†ã‚£ãƒƒã‚¯å›žå¸°ãŒä½¿ã‚ã‚Œã¦ã„ã‚‹ã“ã¨ã«æ°—ã¥ã„ãŸã€‚ ã‚‚ã¡ã‚ん世間ã«ã¯ã‚‚ã£ã¨è¡¨ç¾åŠ›ã®ã‚る分類器ãŒãŸãã•ã‚“ã‚ã‚‹ã‘ã©ã€å•é¡Œã«ã‚ˆã£ã¦ã©ã‚Œãらã„複雑ãªãƒ¢ãƒ‡ãƒ«ãŒé©ã—ã¦ã„ã‚‹ã‹ã¯ç•°ãªã‚‹ã€‚ ãã‚Œã«ã€å„特徴é‡ãŒã©ã®ã‚ˆã†ã«åƒãã‹é‡ã¿ã‹ã‚‰ç¢ºèªã—ãŸã‚Šã€å˜ç´”ãªãƒ¢ãƒ‡ãƒ«ãªã®ã§ã‚¹ã‚³ã‚¢ã‚’ベンãƒãƒžãƒ¼ã‚¯ã¨ã—ã¦åˆ©ç”¨ã™ã‚‹ã€ã¨ã„ã£ãŸç”¨é€”ã‚‚ã‚るらã—ã„。 今回ã¯ã€ãã‚“ãªãƒã‚¸ã‚¹ãƒ†ã‚£ãƒƒã‚¯å›žå¸°ã‚’ scikit-learn ã®å®Ÿè£…ã§è©¦ã—ã¦ã¿ã‚‹ã€‚ 使ã£ãŸç’°å¢ƒã¯æ¬¡ã®é€šã‚Šã€‚ $ sw_vers ProductName: Mac OS X ProductVersion: 10.13.6 BuildVersion: 17G65 $ python -V Python 3.6.6 $ pip list --format=columns | grep -i scikit-learn scikit-learn 0.19.2 下準備 ã¾ãšã¯ scikit-
Pythonã®å¼•æ•°ã¯ã™ã¹ã¦å‚照渡㗠- Qiita
In [27]: i = 1 In [28]: id(i) Out[28]: 140513240953112 In [29]: def func(a): ....: print a, id(a) ....: a += 1 ....: print a, id(a) ....: return a ....: In [30]: b = func(i) 1 140513240953112 <- ã“ã“ã¾ã§ã¯åŒã˜å ´æ‰€(id)ã‚’å‚ç…§ã—ã¦ã„ã‚‹ 2 140513240953088 <- ã“ã“ã§åˆ¥å ´æ‰€(id)ã¸ã‚³ãƒ”ーã•ã‚ŒãŸ In [31]: id(b) Out[31]: 140513240953088 In [32]: i = 1 In [33]: id(i) Out[33]: 140513240953112 In [34]: def func(a): ....: print a, id(a) ....: r
Pythonã§åŸºæœ¬ã®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ を書ã„ã¦ã¿ãŸ - Something Beyond
2015-01-05 Pythonã§åŸºæœ¬ã®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ を書ã„ã¦ã¿ãŸ Programming アルゴリズムをå¦ã¶æ„義ã¿ãŸã„ãªã‚‚ã®ã¯ã„ã‚ã„ã‚ãªã¨ã“ã‚ã§èªžã‚Šå°½ãã•ã‚Œã¦ã„ã‚‹ã¨æ€ã†ã®ã§ç§ã‹ã‚‰ã¯ç‰¹ã«ã‚³ãƒ¡ãƒ³ãƒˆã—ã¾ã›ã‚“ãŒã€ä»Šå›žã®å‹‰å¼·ã«åˆ©ç”¨ã—ãŸæ›¸ç±ã§ã‚‚引用ã•ã‚Œã¦ã„ãŸè¨€è‘‰ãŒå°è±¡çš„ãªã®ã§ã€è¨˜ã—ã¦ãŠãã¾ã™ã€‚ 最先端ã®æ©Ÿæ¢°ã‚’使ã£ã¦è£½å“ã‚’ã¤ãã‚‹ã®ã¯ç°¡å˜ã§ã€ã—ã‹ã‚‚楽ãªã“ã¨ã ãŒã€åŸºæœ¬æŠ€è¡“を固ã‚ã‚‹å‰ã«æ¥½ãªã»ã†ã«æµã‚Œã¦ã„ã£ã¦ã—ã¾ã£ãŸã€‚俺ã®ã‚ˆã†ãªåŸºæœ¬çš„ãªã“ã¨ãŒãã¡ã‚“ã¨ã§ãã‚‹ãƒãƒ¼ãƒ†ã‚¯ãŒã€ä»Šã€æˆ‘ãŒä¸–ã®æ˜¥ã‚’謳æŒã—ã¦ã„ã‚‹ã‚“ã 。 岡野雅行ã•ã‚“ã¨ã„ã†è·äººã•ã‚“ã®è¨€è‘‰ã®ã‚ˆã†ã§ã™ã€‚ãã†ã„ãˆã°éšåˆ†å‰ã«ã“ã‚“ãªè¨˜äº‹ãŒç››ã‚Šä¸ŠãŒã‚Šã¾ã—ãŸã€‚ 今ã™ã辞ã‚ã¦æ¬²ã—ã„ã€ã€ŒRuby on Rails勉強ã—ã¦ã¾ã™ã€ã€ŒCakePHP勉強ã—ã¦ã¾ã™ã€ 最新技術ã ã‘ã§ã¯ãªãã¦ã€ãã®åŸºç¤Žã¨ãªã‚‹æŠ€è¡“ã‚’ã—ã£ã‹ã‚Šç†è§£ã—ãªã‘ã‚Œã°ãƒ€ãƒ¡ã よã¨ã„ã†ã“ã¨ã§ã—ょã†ã€‚ã¨ã„ã†ã“ã¨ã§åŸºæœ¬ã®ã‚¢ãƒ«ã‚´ãƒª
Python pandas データé¸æŠžå‡¦ç†ã‚’ã¡ã‚‡ã£ã¨è©³ã—ã <後編> - StatsFragments
æ¦‚è¦ ã“ã¡ã‚‰ã®ç¶šã。ã“れ㧠pandas ã§ã®ãƒ‡ãƒ¼ã‚¿é¸æŠžã«ã¤ã„ã¦ã¯ã²ã¨ã¾ãšçµ‚ã‚り。 Python pandas データé¸æŠžå‡¦ç†ã‚’ã¡ã‚‡ã£ã¨è©³ã—ã <å‰ç·¨> - StatsFragments Python pandas データé¸æŠžå‡¦ç†ã‚’ã¡ã‚‡ã£ã¨è©³ã—ã <ä¸ç·¨> - StatsFragments サンプルデータã®æº–å‚™ データ㯠å‰ç·¨ã¨åŒã˜ã‚‚ã®ã‚’使ã†ã€‚ãŸã ã—変数åã¯å¤‰ãˆãŸã€‚ import pandas as pd s1 = pd.Series([1, 2, 3], index = ['I1', 'I2', 'I3']) df1 = pd.DataFrame({'C1': [11, 21, 31], 'C2': [12, 22, 32], 'C3': [13, 23, 33]}, index = ['I1', 'I2', 'I3']) s1 # I1 1 # I2 2 # I3 3 # dtype:
Python pandas データé¸æŠžå‡¦ç†ã‚’ã¡ã‚‡ã£ã¨è©³ã—ã <å‰ç·¨> - StatsFragments
æ¦‚è¦ æ›¸ã„ã¦ã„ã¦é•·ããªã£ãŸãŸã‚ã€ã¾ãšå‰ç·¨ã¨ã—㦠pandas 㧠データを行 / 列ã‹ã‚‰é¸æŠžã™ã‚‹æ–¹æ³•ã‚’å°‘ã—詳ã—ã書ã。特ã«ã€å€‹äººçš„ã«ã¯ã‘ã£ã“ã†é‡è¦ã ã¨æ€ã£ã¦ã„ã‚‹ loc 㨠iloc ã«ã¤ã„㦠日本語ã§æ•´ç†ã—ãŸã‚‚ã®ãŒãªã•ãã†ãªã®ã§ã€‚ サンプルデータã®æº–å‚™ import pandas as pd s = pd.Series([1, 2, 3], index = ['I1', 'I2', 'I3']) df = pd.DataFrame({'C1': [11, 21, 31], 'C2': [12, 22, 32], 'C3': [13, 23, 33]}, index = ['I1', 'I2', 'I3']) s # I1 1 # I2 2 # I3 3 # dtype: int64 df # C1 C2 C3 # I1 11 12 13 # I2 21 22 23 # I3 31 32
1
ãŠçŸ¥ã‚‰ã›
ランã‚ング
ランã‚ング
リリースã€éšœå®³æƒ…å ±ãªã©ã®ã‚µãƒ¼ãƒ“スã®ãŠçŸ¥ã‚‰ã›
最新ã®äººæ°—エントリーã®é…ä¿¡
j次ã®ãƒ–ックマーク
kå‰ã®ãƒ–ックマーク
lã‚ã¨ã§èªã‚€
eコメント一覧を開ã
oページを開ã
{{#tags}}- {{label}}
{{/tags}}