Question

我有下表，我试图在每组diffFlag存储桶中获取第一个和最后一个eventTime之间的差值

t:flip (`orderID`sym`eventTimes`side`diffFlag)!(190710 190973 191017 191124 191286 191294 191297 191298 191348 191404;`EURUSD`EURUSD`EURUSD`EURUSD`EURUSD`EURUSD`EURUSD`EURUSD`EURUSD`EURUSD;(02:08:22.674 02:09:05.930;02:18:00.140 02:18:00.140;02:19:31.614 02:24:21.263;02:24:32.710 02:25:26.757;02:30:21.386 02:30:21.386;02:30:36.206 02:30:36.206;02:30:37.607 02:30:37.607;02:30:39.070 02:30:39.070;02:32:06.494 02:34:45.977;02:34:49.426 02:34:49.426);"BSBBSSSSBB";1 2 3 3 4 4 4 4 5 5)

我了解到以下更新语句，但是它未正确汇总。每当diffFlag计数> 1

时，timeDiff应该重置

update sums timeDiff by diffFlag from update timeDiff:(first eventTimes)-last eventTimes by diffFlag from t

任何帮助将不胜感激！

谢谢

Answer 1

q)select {last[x] - first[x]}[last eventTimes] by diffFlag from t
diffFlag| eventTimes  
--------| ------------
1       | 00:00:43.256
2       | 00:00:00.000
3       | 00:00:54.047
4       | 00:00:00.000
5       | 00:00:00.000

更新原始表：

q)update timeDiff:{last[x] - first[x]}[last eventTimes] by diffFlag from t
orderID sym    eventTimes                side diffFlag timeDiff    
-------------------------------------------------------------------
190710  EURUSD 02:08:22.674 02:09:05.930 B    1        00:00:43.256
190973  EURUSD 02:18:00.140 02:18:00.140 S    2        00:00:00.000
191017  EURUSD 02:19:31.614 02:24:21.263 B    3        00:00:54.047
191124  EURUSD 02:24:32.710 02:25:26.757 B    3        00:00:54.047
191286  EURUSD 02:30:21.386 02:30:21.386 S    4        00:00:00.000
191294  EURUSD 02:30:36.206 02:30:36.206 S    4        00:00:00.000
191297  EURUSD 02:30:37.607 02:30:37.607 S    4        00:00:00.000
191298  EURUSD 02:30:39.070 02:30:39.070 S    4        00:00:00.000
191348  EURUSD 02:32:06.494 02:34:45.977 B    5        00:00:00.000
191404  EURUSD 02:34:49.426 02:34:49.426 B    5        00:00:00.000

KDB中第一次和最后一次之间的差异

1 个答案: