SQL: count all records with consecutive occurrence of same value for each device set and return the highest count

问题

I want to find out how many times a particular value occured consecutively for a particular partition and then display the higher count for that partition.

For Example if below is the table:

Device ID        speed             DateTime
--------------------------------------------------
07777778999       34               18-12-2016 17:15
07777778123       15               18-12-2016 18:10
07777778999       34               19-12-2016 19:30
07777778999       34               19-12-2016 12:15
07777778999       20               19-12-2016 13:15
07777778999       20               20-12-2016 11:15
07777778123       15               20-12-2016 9:15
07777778128       44               20-12-2016 17:15
07777778123       15               20-12-2016 17:25
07777778123       12               20-12-2016 17:35
07777778999       34                20-12-2016 17:45
07777778999       34               20-12-2016 17:55
07777778999       34               20-12-2016 18:50
07777778999       34               20-12-2016 18:55

I want to know for each device what is highest number of times the same speed appeared consecutively.

So if i partition them by device id, i would get the belo table

Device ID        speed             DateTime
--------------------------------------------------
07777778999       34               18-12-2016 17:15
07777778999       34               19-12-2016 19:30
07777778999       34               19-12-2016 12:15
07777778999       20               19-12-2016 13:15
07777778999       20               20-12-2016 11:15
07777778999       34                20-12-2016 17:45
07777778999       34               20-12-2016 17:55
07777778999       34               20-12-2016 18:50
07777778999       34               20-12-2016 18:55
07777778123       15               18-12-2016 18:10
07777778123       15               20-12-2016 9:15
07777778123       15               20-12-2016 17:25
07777778123       12               20-12-2016 17:35
07777778128       44               20-12-2016 17:15
-----------------------------------------------------------------

So my required output would be like

Device ID        speed             highcount
--------------------------------------------------
07777778999       34               4
07777778123       15               3

note that 07777778128 did not appear as there were no values which repeated consecutively```

What would be the possible way to achieve this. i was able to get the the count of all consecutive values for each device but then it doesn't give the highest rather gives count of all such consecutive groups

回答1:

This is a form of gaps-and-islands. You can use a difference of row numbers to get the islands:

select device_id, speed, count(*) as num_times
from (select t.*,
             row_number() over (partition by device_id order by datetime) as seqnum,
             row_number() over (partition by device_id, speed order by datetime) as seqnum_s
      from t
     ) t
group by device_id, speed, (seqnum - seqnum_s);

Then, to get the max, use another layer of window functions:

select device_id, speed, num_times
from (select device_id, speed, count(*) as num_times,
             row_number() over (partition by device_id order by count(*) desc) as seqnum
      from (select t.*,
                   row_number() over (partition by device_id order by datetime) as seqnum,
                   row_number() over (partition by device_id, speed order by datetime) as seqnum_s
            from t
           ) t
      group by device_id, speed, (seqnum - seqnum_s)
     ) ds
where seqnum = 1;

来源：https://stackoverflow.com/questions/59037783/sql-count-all-records-with-consecutive-occurrence-of-same-value-for-each-device

标签

mysql

sql

aggregation

data-analysis