[問題] sapply應用在整個data.fram
[問題類型]:
效能諮詢(我想讓R 跑更快)
[問題敘述]:
最近在用cross-validation來model selection
在目前的樣本隨機抽取(with replacement)N筆後配適模型
上述步驟重複100次,
接著N改為N+5,再重複100次,如此N+5k一直做下去
目前是用for-loop
但覺得跑得好慢
於是在想說不知是否能用apply族的函數
但看了說明,大多是對data.fram的每個欄執行function
好像沒發現有對整個data.fram執行function
故上來請教一下
有想過不然創建個維度100的LIST,每個LIST都是一整個data.fram
不過還沒試不知道可不可行= =
[程式範例]:
http://pastebin.com/avAHvyhd
[環境敘述]:
R version 3.2.2 (2015-08-14)
Platform: x86_64-w64-mingw32/x64 (64-bit)
Running under: Windows 8 x64 (build 9200)
[關鍵字]:
選擇性,也許未來有用
--
○ 啊~ ○○ 咿~ ○ 喔~ 嗯~
ㄥ︹○ ︽〕 ○\︿ ○
》 ﹨︿ ○︿/︿ ○≦\_
(十面埋伏) (針鋒相對) (腹背受敵) (騰雲駕霧)
--
※ 發信站: 批踢踢實業坊(ptt.cc), 來自: 184.6.253.131
※ 文章網址: https://www.ptt.cc/bbs/R_Language/M.1443616001.A.654.html
→
09/30 20:36, , 1F
09/30 20:36, 1F
→
09/30 20:43, , 2F
09/30 20:43, 2F
→
09/30 20:43, , 3F
09/30 20:43, 3F
→
09/30 20:44, , 4F
09/30 20:44, 4F
→
09/30 20:47, , 5F
09/30 20:47, 5F
→
09/30 20:47, , 6F
09/30 20:47, 6F
→
09/30 20:54, , 7F
09/30 20:54, 7F
→
09/30 21:11, , 8F
09/30 21:11, 8F
→
09/30 21:12, , 9F
09/30 21:12, 9F
→
09/30 21:13, , 10F
09/30 21:13, 10F
→
09/30 21:14, , 11F
09/30 21:14, 11F
→
09/30 21:14, , 12F
09/30 21:14, 12F
→
09/30 21:26, , 13F
09/30 21:26, 13F
→
09/30 21:26, , 14F
09/30 21:26, 14F
data fram的dim()是428*23, 全都是numerical
→
09/30 22:56, , 15F
09/30 22:56, 15F
→
09/30 23:11, , 16F
09/30 23:11, 16F
推
09/30 23:12, , 17F
09/30 23:12, 17F
→
09/30 23:12, , 18F
09/30 23:12, 18F
推
09/30 23:12, , 19F
09/30 23:12, 19F
→
09/30 23:12, , 20F
09/30 23:12, 20F
→
09/30 23:14, , 21F
09/30 23:14, 21F
→
09/30 23:15, , 22F
09/30 23:15, 22F
第2點有做到
其餘三點等等來研究研究,晚點若研究出來就來更新一下XD
謝謝兩位建議的方向!
※ 編輯: MADNUG (184.6.253.131), 10/01/2015 00:58:40
→
10/01 06:48, , 23F
10/01 06:48, 23F
推
10/01 10:37, , 24F
10/01 10:37, 24F
→
10/01 10:38, , 25F
10/01 10:38, 25F
→
10/01 10:39, , 26F
10/01 10:39, 26F
→
10/01 10:40, , 27F
10/01 10:40, 27F
→
10/01 12:43, , 28F
10/01 12:43, 28F
→
10/01 12:43, , 29F
10/01 12:43, 29F
→
10/01 13:17, , 30F
10/01 13:17, 30F
→
10/01 13:17, , 31F
10/01 13:17, 31F
→
10/01 13:19, , 32F
10/01 13:19, 32F
→
10/01 16:54, , 33F
10/01 16:54, 33F
→
10/01 16:54, , 34F
10/01 16:54, 34F
→
10/01 16:56, , 35F
10/01 16:56, 35F
→
10/01 16:58, , 36F
10/01 16:58, 36F
→
10/01 20:58, , 37F
10/01 20:58, 37F
→
10/01 21:05, , 38F
10/01 21:05, 38F
→
10/01 21:05, , 39F
10/01 21:05, 39F
→
10/01 21:16, , 40F
10/01 21:16, 40F
→
10/01 21:33, , 41F
10/01 21:33, 41F
→
10/01 21:37, , 42F
10/01 21:37, 42F