[B! compression] somemoã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯

somemo id:somemo

compressionã«é–¢ã™ã‚‹somemoã®ãƒ–ãƒƒã‚¯ãƒžãƒ¼ã‚¯ (10)

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

${{author_name}}$
{{author_name}}{{created}}
{{ #comment }}{{ comment }}{{ /comment }}
- {{ label }}

${{author_name}}$

{{{comment_expanded}}}

{{label}}

{{#is_bookmark}}ãƒªã‚¹ãƒˆ{{/is_bookmark}}{{^is_bookmark}}ãƒªãƒ³ã‚¯{{/is_bookmark}}

DO++ : é€éŽçš„ãƒ‡ãƒ¼ã‚¿åœ§ç¸®
å¯é€†ãƒ‡ãƒ¼ã‚¿åœ§ç¸®åˆ†é‡Žã§ã€ç¾åœ¨ç ”ç©¶ãŒç››ã‚“ãªåˆ†é‡Žã®ä¸€ã¤ãŒã€ãƒ‡ãƒ¼ã‚¿ã‚’åœ§ç¸®ã—ãŸçŠ¶æ…‹ã®ã¾ã¾å®šæ•°æ™‚é–“ã§ãƒ©ãƒ³ãƒ€ãƒ ã‚¢ã‚¯ã‚»ã‚¹ã‚’ã‚µãƒãƒ¼ãƒˆã™ã‚‹ãƒ‡ãƒ¼ã‚¿åœ§ç¸®æ–¹å¼ã§ã™ï¼ˆword RAMãƒ¢ãƒ‡ãƒ«ã§O(log n)ã‚µã‚¤ã‚ºã®å¾©å…ƒãŒå®šæ•°æ™‚é–“)ã€‚ ã“ã‚Œã¯ã€ãƒ‡ãƒ¼ã‚¿ã‚’ã‚ãŸã‹ã‚‚åœ§ç¸®ã—ã¦ã„ãªã„ã‹ã®ã‚ˆã†ã«æ‰±ãˆã‚‹ãŸã‚ã€é€éŽçš„ãƒ‡ãƒ¼ã‚¿åœ§ç¸®/æ§‹é€ ã¨å‘¼ã°ã‚Œã¦ã„ã¾ã™ï¼ˆè‹±èªžã ã¨ã¾ã æ±ºã¾ã£ã¦ãªã„ï¼Ÿ)ã€‚ ä¾‹ãˆã°1GBã®ãƒ‡ãƒ¼ã‚¿ã‚’åœ§ç¸®ã—ãŸçŠ¶æ…‹ã§ã€é€”ä¸300MBç›®ã‹ã‚‰4Byteã ã‘å¾©å…ƒã—ã‚ˆã†ã¨ã„ã†ã®ãŒå®šæ•°æ™‚é–“ã§å®Ÿç¾ã§ãã‚‹ã‚ã‘ã§ã™ã€‚ã“ã‚Œã¯ç†è«–çš„ã«ã‚‚ã‹ãªã‚Šå¼·ã„ã“ã¨ã‚’ã„ã£ã¦ã„ã¦ï¼Œä¾‹ãˆã°ä»Šã‚ã‚‹ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã‚„ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ãŒã€O(T)æ™‚é–“ã§ã‚ã‚‹å•é¡Œã‚’è§£ã‘ã‚‹ã¨ã„ã†ã®ãŒã‚ã£ãŸã‚‰ã€ãã‚Œã‚’å…¨ãåŒã˜O(T)æ™‚é–“ã®ã¾ã¾ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã‚’åœ§ç¸®ã—ä½œæ¥é ˜åŸŸé‡ã‚’æ¸›ã‚‰ã™ã“ã¨ãŒã§ãã¾ã™ (ä¸€å¿œãƒ‡ãƒ¼ã‚¿æ§‹é€ ã«å¯¾ã—èªã¿è¾¼ã¿æ“ä½œã—ã‹ç„¡ã„å ´åˆã€‚æ›¸ãè¾¼ã¿ã‚‚ã‚ã‚‹å ´åˆã¯ã¾ãŸã¡ã‚‡ã£ã¨é¢å€’ã«ãªã‚‹ï¼‰ ã“ã®ãƒ‡ãƒ¼ã‚¿ã‚’åœ§ç¸®ã—ãŸã¾ã¾æ‰±ã†
somemo 2015/07/30
compression
ãƒªãƒ³ã‚¯
ã€Œé«˜é€Ÿæ–‡å—åˆ—è§£æžã®ä¸–ç•Œã€ã‚’èªã‚€ã®ã«å‚è€ƒã«ã—ãŸè³‡æ–™
ç¾æ™‚ç‚¹ã§ã¯ã€ãƒ“ãƒƒã‚°ãƒ‡ãƒ¼ã‚¿ã§ã¯ãªãã€ã‚µãƒ³ãƒ—ãƒªãƒ³ã‚°ã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ã‚’æ‰±ã£ã¦ã„ã‚‹ã®ã§ã™ãŒã€ã€Œå¤§è¦æ¨¡ã€ãƒ‡ãƒ¼ã‚¿ã§ã‚ã‚‹ã“ã¨ã«ã¯é•ã„ãªãã€ãã‚Œã‚‰ã‚’å¦‚ä½•ã«é«˜é€Ÿã«å‡¦ç†ã™ã‚‹ã‹ã«é ã‚’æ‚©ã¾ã›ã¦ã„ã¾ã™ã€‚ ã“ã®æœ¬ã¯ã€ãã®æ‚©ã¿ã‚’è§£æ±ºã™ã‚‹åŠ©ã‘ã«ãªã‚Šã¾ã—ãŸã€‚ å®Ÿéš›ã«ã¯ã€ã„ãã‚‰ã‹ã®å‰æçŸ¥è˜ãŒå¿…è¦ã§ã™ã—ã€å¯¾è±¡ã¨ã™ã‚‹å•é¡Œã‚„é–‹ç™ºç’°å¢ƒã‚‚è€ƒæ…®ã—ãªã‘ã‚Œã°ãªã‚Šã¾ã›ã‚“ã§ã—ãŸã€‚ éƒ¨åˆ†çš„ãªèª¿æ•´ã—ã‹ã§ãã¦ã„ã¾ã›ã‚“ãŒã€ãã‚Œã§ã‚‚ã€å½“åˆã¯è‡´å‘½çš„ã¨ã‚‚æ€ã‚ã‚ŒãŸå®Ÿè¡Œé€Ÿåº¦ã‚’ã€ä½•ã¨ã‹ç¾å®Ÿçš„ãªé€Ÿåº¦ã¾ã§æŒã¦ãã‚‹ã“ã¨ãŒã§ãã¾ã—ãŸã€‚ ä»¥ä¸‹ã«ã€ã“ã®æœ¬ã‚’èªã‚€éš›ã«å‚è€ƒã«ã—ãŸãƒšãƒ¼ã‚¸ã€å„ç¨®å®Ÿè£…é–¢é€£ã®ãƒšãƒ¼ã‚¸ãªã©ã¸ã®ãƒªãƒ³ã‚¯ã‚’ï¼ˆå‚™å¿˜éŒ²ã‚‚å…¼ãã¦ï¼‰è¨˜ã—ã¦ãŠãã¾ã™ã€‚ ãªãŠã€ã€Œå‚è€ƒã«ã—ãŸã€ãƒšãƒ¼ã‚¸ã§ã™ã®ã§ã€å¿…ãšã—ã‚‚æœ¬æ›¸ã«ã¦è§£èª¬ã•ã‚Œã¦ã„ã‚‹é …ç›®ã¨ã¯é™ã‚Šã¾ã›ã‚“ã€‚ â€»å®Ÿè£…ã‚³ãƒ¼ãƒ‰ã®é …ç›®ã«è¨˜è¼‰ã—ãŸãƒ©ã‚¤ã‚»ãƒ³ã‚¹ã¯èª¿æŸ»æ™‚ã®ã‚‚ã®ã§ã™ã—ã€æœªç¢ºèªãƒ»æœªè¨˜è¼‰ã®ã‚‚ã®ã‚‚ã‚ã‚Šã¾ã™ã€‚åˆ©ç”¨ã«éš›ã—ã¦ã¯ã€ãã‚Œãžã‚Œç¢ºèªãŒå¿…è¦ã§ã™ã€‚ â– å…¨èˆ¬
somemo 2014/07/27
data structure

algorithm

compression
ãƒªãƒ³ã‚¯
Î³ç¬¦å·ã€Î´ç¬¦å·ã€ã‚´ãƒãƒ ç¬¦å·ã«ã‚ˆã‚‹åœ§ç¸®åŠ¹æžœ - naoyaã®ã¯ã¦ãªãƒ€ã‚¤ã‚¢ãƒªãƒ¼
é€šå¸¸ã®æ•´æ•°ã¯ 32 ãƒ“ãƒƒãƒˆã¯ 4 ãƒã‚¤ãƒˆã®å›ºå®šé•·ã«ã‚ˆã‚‹ãƒã‚¤ãƒŠãƒªç¬¦å·ã§ã™ãŒã€å°ã•ãªæ•°å—ãŒãŸãã•ã‚“å‡ºç¾ã—ã€å¤§ããªæ•°å—ã¯ã»ã¨ã‚“ã©å‡ºç¾ã—ãªã„ã¨ã„ã†ç¢ºçŽ‡åˆ†å¸ƒã®ã‚‚ã¨ã§ã¯ç„¡é§„ãªãƒ“ãƒƒãƒˆãŒç›®ç«‹ã¡ã¾ã™ã€‚ Variable Byte Code (Byte Aligned ç¬¦å·ã¨ã‚‚å‘¼ã°ã‚Œã¾ã™) ã¯æ•´æ•°ã®ç¬¦å·åŒ–æ‰‹æ³•ã®ä¸€ã¤ã§ã€ã“ã®ç„¡é§„ã‚’å¹¾åˆ†è§£æ¶ˆã—ã¾ã™ã€‚è©³ã—ãã¯ Introduction to Information Retrieval (ä»¥ä¸‹ IIR) ã®ç¬¬5ç« ã«æŽ²è¼‰ã•ã‚Œã¦ã„ã¾ã™ã€‚(http://nlp.stanford.edu/IR-book/html/htmledition/variable-byte-codes-1.html ã§å…¬é–‹ã•ã‚Œã¦ã„ã¾ã™) Variable Byte Code ã¯ãã®åã®é€šã‚Šãƒã‚¤ãƒˆãƒ¬ãƒ™ãƒ«ã®å¯å¤‰é•·ç¬¦å·ã§ã€1ãƒã‚¤ãƒˆã®å…ˆé 1ãƒ“ãƒƒãƒˆã‚’ continuation ãƒ“ãƒƒãƒˆã¨ã—ã¦æ‰±ã„ã€ç¶šã 7 ãƒ“ãƒƒãƒˆ
somemo 2014/06/14
compression

algorithm
ãƒªãƒ³ã‚¯
Variable byte codes
Variable byte (VB) encoding uses an integral number of bytes to encode a gap. The last 7 bits of a byte are ``payload'' and encode part of the gap. The first bit of the byte is a continuation bit . It is set to 1 for the last byte of the encoded gap and to 0 otherwise. To decode a variable byte code, we read a sequence of bytes with continuation bit 0 terminated by a byte with continuation bit 1.
somemo 2014/06/14
algorithm

compression
ãƒªãƒ³ã‚¯
Perlã§åœ§ç¸®
5. ãƒ‡ã‚£ã‚¹ã‚¯ãƒªãƒ¼ãƒ‰ã¨åœ§ç¸® HDD vs ãƒ¡ãƒ¢ãƒª -> 10 -6 ~ 10 -9 ã®å·® HDD ä¸Šã®ãƒ‡ãƒ¼ã‚¿ã®ã‚·ãƒ¼ã‚¯ ï½¥ï½¥ï½¥ ãƒ‡ã‚£ã‚¹ã‚¯ã®å›žè»¢å¾…ã¡ã€ãƒ˜ãƒƒãƒ‰ã®ç§»å‹•ã« ms âˆ´ ãƒ‡ã‚£ã‚¹ã‚¯ã‚·ãƒ¼ã‚¯å›žæ•°ã‚’æœ€å°åŒ–ã—ãŸã„ OS ( ã‚«ãƒ¼ãƒãƒ« ) ãƒ–ãƒãƒƒã‚¯å˜ä½ã§ã®èªã¿å‡ºã— é€£ç¶šã—ãŸãƒ‡ãƒ¼ã‚¿ã¯è¿‘ã„å ´æ‰€ã«ä¿å˜ ã‚¢ãƒ—ãƒªã‚±ãƒ¼ã‚·ãƒ§ãƒ³ B+ æœ¨ãªã©æœ€é©ãªãƒ‡ãƒ¼ã‚¿æ§‹é€ ã®é¸æŠž åœ§ç¸®ã—ã¦ä¿å˜ -> æ¶ˆè²»ãƒ–ãƒãƒƒã‚¯æ•°ã‚’æ¸›ã‚‰ã™ 6. å¾©å·å‡¦ç†ã¨è² è· åœ§ç¸®ã—ã¦ä¿å˜ -> CPUå‡¦ç†ã¨ I/O ã®ã‚ªãƒ¼ãƒãƒ¼ãƒ©ãƒƒãƒ—ã§ã‚¹ãƒ«ãƒ¼ãƒ—ãƒƒãƒˆå‘ä¸Š åœ§ç¸®æ¸ˆã¿ãƒ‡ãƒ¼ã‚¿ã®èªã¿å‡ºã— ï½¥ï½¥ï½¥ I/O è² è· åœ§ç¸®ã•ã‚ŒãŸãƒ‡ãƒ¼ã‚¿ã®å±•é–‹ ï½¥ï½¥ï½¥ CPU è² è· ãƒ‡ãƒ¼ã‚¿ãŒå°ã•ã„ ï½¥ï½¥ï½¥ ã‚ãƒ£ãƒƒã‚·ãƒ¥ã«è¼‰ã‚Šã‚„ã™ã„
somemo 2014/06/14
Compression

algorithm
ãƒªãƒ³ã‚¯
MySQLã®æ–°ã—ã„InnoDB ãƒšãƒ¼ã‚¸I/Oåœ§ç¸®æ©Ÿèƒ½ã«ã¤ã„ã¦è§£æžã—ã¦ã¿ãŸ: hasegaw blog
InnoDBã«ã¯ãƒ‡ãƒ¼ã‚¿ã®åœ§ç¸®æ©Ÿèƒ½ãŒã‚ã‚Šã¾ã™ãŒã€ãƒ‘ãƒ•ã‚©ãƒ¼ãƒžãƒ³ã‚¹ãŒä½Žã„ã“ã¨ã‹ã‚‰ã‚ã¾ã‚Šä½¿ã‚ã‚Œã¦ã„ã¾ã›ã‚“ã€‚ãŸã ä»Šå¹´ã® Percona Live ã§ Oracle MySQL, MariaDB, ãã—ã¦ Percona Server ãŒæ–°ã—ã„ InnoDB Compression ã‚’å‡ºã—ã¦ãã¾ã—ãŸã€‚ã“ã‚Œã¯Fusion-ioã® R&D ãƒãƒ¼ãƒ ãŒãƒ•ãƒ©ãƒƒã‚·ãƒ¥ã‚¹ãƒˆãƒ¬ãƒ¼ã‚¸å‘ã‘ã® MySQL é«˜é€ŸåŒ–ã®ä¸€ç’°ã§é–‹ç™ºã—ãŸãƒ‘ãƒƒãƒãŒå…ƒã«ãªã£ã¦ã„ã¾ã™ã€‚ã¡ãªã¿ã«ç§ã¯ Fusion-io ã®ç¤¾å“¡ã§ã™ã®ã§ã“ã®ç™ºè¡¨ã‚’ãƒ¯ã‚¯ãƒ†ã‚«ã—ã¦å¾…ã£ã¦ã„ãŸã®ã§ã™ãŒã€æŠ˜è§’ã‚³ãƒ¼ãƒ‰ãŒä¸€èˆ¬ã«ãƒªãƒªãƒ¼ã‚¹ã•ã‚ŒãŸã®ã§ã€ã‚½ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‰ã‚’çœºã‚ã¦å‹•ä½œã‚’èª¿ã¹ã‚‹ã“ã¨ã«ã—ã¾ã—ãŸã€‚ å‚è€ƒã«ã—ãŸã®ã¯ MySQL Server Snapshots (labs.mysql.com) ã«ã‚ã‚‹ MySQL with InnoDB PageIO Compression ã®ã‚½ãƒ¼ã‚¹ã‚³ãƒ¼ãƒ‰ã€ãŠã‚ˆã³
somemo 2014/04/03
MySQL

compression

MariaDB
ãƒªãƒ³ã‚¯
ãƒ†ã‚ã‚¹ãƒˆåœ§ç¸®ã¯ã“ã‚Œä¸€å†Šã§OKï¼ï¼Ÿãªå„ªè‰¯æ›¸ç±ã€ŒThe Burrows-Wheeler Transformã€ã‚’èªã‚“ã - EchizenBlog-Zwei
ä»¥å‰ã‚ˆã‚Šæ°—ã«ãªã£ã¦ã„ãŸæ›¸ç±ã€ŒThe Burrows-Wheeler Transf orm Data Compression, Suffix Arrays, and Pattern matchingã€ã‚’èªã‚€æ©Ÿä¼šã‚’å¾—ã‚‹ã“ã¨ãŒã§ããŸã€‚ãã‚Œãªã‚Šã«é«˜é¡ãªæœ¬ã ã£ãŸã®ã§è³¼å…¥ãŒèºŠèº‡ã£ã¦ã„ãŸã®ã ã‘ã©ã€ã“ã‚Œã¯è‡ªåˆ†ç”¨ã«è³¼å…¥ã—ã¦ã‚‚ã„ã„ã‹ã‚‚ã€‚ã¨ã„ã†ãã‚‰ã„ã®è‰¯æ›¸ã ã£ãŸã®ã§ç´¹ä»‹ã—ã¦ãŠãã€‚ æœ¬æ›¸ã¯ã‚¿ã‚¤ãƒˆãƒ«ã®ã¨ãŠã‚ŠBWT(Burrows-Wheelerå¤‰æ›)ã«é–¢ã™ã‚‹æ›¸ç±ã€‚ã‚µãƒ–ã‚¿ã‚¤ãƒˆãƒ«ã«ã‚ã‚‹ã‚ˆã†ã«ãƒ‡ãƒ¼ã‚¿åœ§ç¸®ã‚„SuffixArrayã«ã‚ˆã‚‹å…¨æ–‡æ¤œç´¢ã«ã¤ã„ã¦ã‚‚å……å®Ÿã—ãŸå†…å®¹ã«ãªã£ã¦ã„ã‚‹ã€‚æœ€å¾Œã®Pattern matchingã¯ãƒ†ã‚ã‚¹ãƒˆã‹ã‚‰æ¤œç´¢ã‚ãƒ¼ã¨exactã«ãƒžãƒƒãƒã—ãŸã€ã‚‚ã—ãã¯é¡žä¼¼ã—ãŸç®‡æ‰€ã‚’å–ã‚Šå‡ºã™ã‚ˆã€ã¨ã„ã†è©±ã€‚2008å¹´ã®æœ¬ãªã®ã§æ¯”è¼ƒçš„æ–°ã—ã„è©±é¡Œã‚‚æ‰±ã£ã¦ã„ã¦æº€è¶³åº¦ãŒé«˜ã„ã€‚ ã¾ãŸæœ¬æ›¸ã®ç‰¹è‰²ã¯åœ§ç¸®ã‚ã‚Šãã§å§‹ã¾ã‚Šã€ãã“ã‹ã‚‰å…¨æ–‡æ¤œç´¢å¯èƒ½ãª
somemo 2013/12/27
compression

algorithm
ãƒªãƒ³ã‚¯
é€£æƒ³é…åˆ—ã¯ãƒˆãƒ©ã‚¤ã§ã—ã‚‡çš„ãªè©±ãŒã§ã¦ã„ãŸã®ã§å…¥é–€è¨˜äº‹ã‚’æ›¸ã„ã¦ã¿ãŸ - EchizenBlog-Zwei
ãªã«ã‚„ã‚‰Dan Kogaiæ°ã®ä»¥ä¸‹ã®è¨˜äº‹ãŒè©±é¡Œã«ãªã£ã¦ã„ã‚‹æ§˜åã€‚ 404 Blog Not Found:Algorithm - é€£æƒ³é…åˆ—ã®å®Ÿè£…ã¨ã—ã¦ã®ãƒãƒƒã‚·ãƒ¥ã¯ã‚ªãƒ¯ã‚³ãƒ³? é€£æƒ³é…åˆ—(ã‚ãƒ¼ãƒ¯ãƒ¼ãƒ‰ã‚’æŠ•ã’ã‚‹ã¨å¯¾å¿œã™ã‚‹å€¤ãŒè¿”ã£ã¦ãã‚‹ãƒ‡ãƒ¼ã‚¿æ§‹é€ )ã¯ãƒãƒƒã‚·ãƒ¥ãƒ†ãƒ¼ãƒ–ãƒ«ã§å®Ÿè£…ã™ã‚‹ã®ã§ã¯ãªãã€ã“ã‚Œã‹ã‚‰ã¯ãƒˆãƒ©ã‚¤(trie)æœ¨ã‚’ä½¿ã†ã®ãŒã‚¤ã‚±ã¦ã‚‹ï¼(æ„è¨³)ã¨ã„ã†å†…å®¹ã ã£ãŸã€‚ é€£æƒ³é…åˆ—ã«ãƒãƒƒã‚·ãƒ¥ãƒ†ãƒ¼ãƒ–ãƒ«ã‚’ä½¿ã†ã®ãŒè‰¯ã„ã‹æ‚ªã„ã‹ã«ã¤ã„ã¦ã¯è‰²ã€…ã¨æ„è¦‹ãŒã‚ã‚‹ã¨æ€ã†ã®ã§ç‰¹ã«ã“ã®è¨˜äº‹ã§ã¯è§¦ã‚Œãªã„ã€‚ ä»Šå›žã¯é€£æƒ³é…åˆ—ã¨ã—ã¦ä½¿ãˆã‚‹ã¨è©±é¡Œã®ãƒˆãƒ©ã‚¤æœ¨ã¨ã¯ãªã‚“ãžã€ã¨ã„ã†å…¥é–€çš„ãªè¨˜äº‹ã«ã—ã‚ˆã†ã¨æ€ã†ã€‚ ãƒˆãƒ©ã‚¤æœ¨ãŒæŒã¤æ©Ÿèƒ½ æœ€åˆã«ãƒˆãƒ©ã‚¤ãŒæŒã¤ä»¥ä¸‹ã®3ã¤ã®æ©Ÿèƒ½ã«ã¤ã„ã¦èª¬æ˜Žã™ã‚‹ã€‚ - lookup - common-prefix-search - predictive-searchã¾ãšãƒˆãƒ©ã‚¤ã¯é€£æƒ³é…åˆ—ã¨ã—ã¦åˆ©ç”¨ã§ãã‚‹ã€‚ã¤ã¾ã‚Šã‚ãƒ¼ãƒ¯ãƒ¼ãƒ‰ã¨å€¤ã®ãƒšã‚¢ã‚’ç™»
somemo 2013/12/27
Algorithm

DoubleArray

compression
ãƒªãƒ³ã‚¯
Googleã®æ–°ã—ã„åœ§ç¸®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ Zopfliã«ã¤ã„ã¦èª¿ã¹ãŸã€‚ - EchizenBlog-Zwei
Googleã®æ–°ã—ã„åœ§ç¸®ã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ Zopfliã«ã¤ã„ã¦èª¿ã¹ãŸã®ã§ãƒ¡ãƒ¢ã—ã¦ãŠãã€‚ Compress data more densely with Zopfli - Google Developers Blog deflateã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ zopfliã¯deflateã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã«åŸºã¥ã„ãŸåœ§ç¸®ãƒ©ã‚¤ãƒ–ãƒ©ãƒªã€‚deflateã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã¯ãƒ‡ãƒ¼ã‚¿ã‚’LZSSã¨ã„ã†LZ77ã‚’æ”¹è‰¯ã—ãŸåœ§ç¸®æ³•ã§åœ§ç¸®ã—ã€ãã®å¾Œãƒãƒ•ãƒžãƒ³ç¬¦å·ã§ç¬¦å·åŒ–ã—ãŸã‚‚ã®ã€‚ deflateã‚¢ãƒ«ã‚´ãƒªã‚ºãƒ ã®å®Ÿè£…ã¨ã—ã¦ã¯zlibã‚„gzipãŒã‚ã‚‹ã€‚zlibã¨gzipã®é•ã„ã¯ãƒ˜ãƒƒãƒ€ã¨ãƒ•ãƒƒã‚¿ã®æŒã£ã¦ã„ã‚‹æƒ…å ±ã€‚ ä½¿ã£ã¦ã¿ã‚‹ googlecode(http://code.google.com/p/zopfli/)ã‹ã‚‰å–ã£ã¦ãã‚‹ã€‚ $$ git clone https://code.google.com/p/zopfli/ $$ cd zopfli $$ ma
somemo 2013/12/27
algorithm

compression
ãƒªãƒ³ã‚¯
ã€Œé«˜é€Ÿæ–‡å—åˆ—è§£æžã®ä¸–ç•Œã€ã‚’èªã‚€å‰ã«çŸ¥ã£ã¦ãŠãã¨è‰¯ã„ã“ã¨ - EchizenBlog-Zwei
ã€Œé«˜é€Ÿæ–‡å—åˆ—è§£æžã®ä¸–ç•Œã€ã¨ã„ã†å¤§å¤‰ã™ã°ã‚‰ã—ã„æœ¬ãŒç™ºå£²ã•ã‚ŒãŸã€‚ã‚ã‚Šã¨æ•·å±…ãŒé«˜ã„æœ¬ã§ã¯ã‚ã‚‹ã®ã§èªã‚€å‰ã«çŸ¥ã£ã¦ãŠãã¨ã‚ˆã•ãã†ãªã“ã¨ã‚’æ›¸ã„ã¦ãŠãã€‚ ã€Œé«˜é€Ÿæ–‡å—åˆ—è§£æžã€ã¨ã¯ æœ¬æ›¸ã§ã„ã†é«˜é€Ÿæ–‡å—åˆ—è§£æžã¨ã„ã†ã®ã¯ä¸»ã«2ã¤ã®ã“ã¨ã‚’æŒ‡ã—ã¦ã„ã‚‹ã€‚ã²ã¨ã¤ã¯ãƒ‡ãƒ¼ã‚¿ã‚’åœ§ç¸®ã—ã¦å°ã•ãã—ã¦ãƒ‡ã‚£ã‚¹ã‚¯ã‚ˆã‚Šãƒ¡ãƒ¢ãƒªã€ãƒ¡ãƒ¢ãƒªã‚ˆã‚Šã‚ãƒ£ãƒƒã‚·ãƒ¥ã¨ã„ã†ã‚ˆã†ã«ã‚ˆã‚Šé«˜é€Ÿãªè¨˜æ†¶è£…ç½®ã§æ‰±ã„ã¾ã—ã‚‡ã†ã€ã¨ã„ã†è©±ã€‚ã‚‚ã†ã²ã¨ã¤ã¯ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã‚’å·¥å¤«ã™ã‚‹ã“ã¨ã§è¤‡é›‘ãªæ“ä½œã‚‚ãã“ãã“é«˜é€Ÿã«æ‰±ãˆã¾ã™ã‚ˆã€ã¨ã„ã†è©±ã€‚ã¤ã¾ã‚Šã€Œåœ§ç¸®ã€ã®è©±ã¨ã€ŒåŠ¹çŽ‡çš„ãªãƒ‡ãƒ¼ã‚¿æ§‹é€ ã€ã®è©±ãŒã‚ã‚‹ã¨è€ƒãˆã¦ãŠãã¨è‰¯ã„ã€‚ ã‚ãƒ¼ãƒ¯ãƒ¼ãƒ‰ã¯3ã¤ ã‚ªãƒ“ã«ã‚‚æ›¸ã„ã¦ã‚ã‚‹ã‘ã‚Œã©ã€æœ¬æ›¸ãŒä¸»ã«æ‰±ã†ã®ã¯ã€ŒBWTã€ã€Œç°¡æ½”ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã€ã€Œã‚¦ã‚§ãƒ¼ãƒ–ãƒ¬ãƒƒãƒˆæœ¨ã€ã®3ã¤ã€‚å…·ä½“çš„ã«ã¯ã€ŒBWTã€ãŒã€Œåœ§ç¸®ã€ã«é–¢ã‚ã£ã¦ã„ã¦ã€Œã‚¦ã‚§ãƒ¼ãƒ–ãƒ¬ãƒƒãƒˆæœ¨ã€ãŒã€ŒåŠ¹çŽ‡çš„ãªãƒ‡ãƒ¼ã‚¿æ§‹é€ ã€ã«é–¢ã‚ã£ã¦ã„ã‚‹ã€‚ã€Œç°¡æ½”ãƒ‡ãƒ¼ã‚¿æ§‹é€ ã€ã¯åŸºæœ¬çš„ãªé“å…·ã¨ã—ã¦æœ¬æ›¸ã®è‰²ã€…ãªã¨ã“ã‚ã§å‡ºã¦
somemo 2013/01/09
data structure

compression

NLP

algorithm
ãƒªãƒ³ã‚¯
1