CLICKHOUSE GROUP BY 4 毫秒

Question

我想平均 4 个连续的行（每 4 毫秒）。我想 GROUP BY N 毫秒。如何使用时间戳或使用模数的 UNIX 时间戳来做到这一点？基本上，我想按 4 行分组，然后平均这 4 行，然后移动到接下来的 4 行。请分享脚本。 Im stuck here

INSERT INTO sensor_1000 Values
('2021-01-01 00:00:00.000', 1.52), ('2021-01-01 00:00:00.001', 1.54), ('2021-01-01 00:00:00.002', 1.42), ('2021-01-01 00:00:00.003', 1.54), ('2021-01-01 00:00:00.004', 1.42), ('2021-01-01 00:00:00.005', 1.52), ('2021-01-01 00:00:00.006', 1.54), ('2021-01-01 00:00:00.007', 1.42), ('2021-01-01 00:00:00.008', 1.54), ('2021-01-01 00:00:00.009', 1.42),
('2021-01-01 00:00:00.010', 1.55), ('2021-01-01 00:00:00.011', 1.45), ('2021-01-01 00:00:00.012', 1.55), ('2021-01-01 00:00:00.013', 1.45), ('2021-01-01 00:00:00.014', 1.35), ('2021-01-01 00:00:00.015', 1.55), ('2021-01-01 00:00:00.016', 1.45), ('2021-01-01 00:00:00.017', 1.55), ('2021-01-01 00:00:00.018', 1.45), ('2021-01-01 00:00:00.019', 1.35),
('2021-01-01 00:00:00.020', 1.54), ('2021-01-01 00:00:00.021', 1.44), ('2021-01-01 00:00:00.022', 1.54), ('2021-01-01 00:00:00.023', 1.44), ('2021-01-01 00:00:00.024', 1.34), ('2021-01-01 00:00:00.025', 1.54), ('2021-01-01 00:00:00.026', 1.44), ('2021-01-01 00:00:00.027', 1.54), ('2021-01-01 00:00:00.028', 1.44), ('2021-01-01 00:00:00.029', 1.34),
('2021-01-01 00:00:00.030', 1.53), ('2021-01-01 00:00:00.031', 1.43), ('2021-01-01 00:00:00.032', 1.53), ('2021-01-01 00:00:00.033', 1.43), ('2021-01-01 00:00:00.034', 1.33), ('2021-01-01 00:00:00.035', 1.53), ('2021-01-01 00:00:00.036', 1.43), ('2021-01-01 00:00:00.037', 1.53), ('2021-01-01 00:00:00.038', 1.43), ('2021-01-01 00:00:00.039', 1.33),

Answer 1

SELECT
    toUnixTimestamp64Milli(now64(3)) AS x,
    intDiv(x, 4) * 4 AS y,
    x - y AS diff
┌─────────────x─┬─────────────y─┬─diff─┐
│ 1640362330078 │ 1640362330076 │    2 │
└───────────────┴───────────────┴──────┘


select intDiv( toUnixTimestamp64Milli ( ts ), 4) * 4 rounded_ts, ....
from
group by rounded_ts