PostgreSQL的範圍小時計數明智數據

I have simple table 

    discount_code | updated_date 
    ---------------------------- 
    L1   | 2017-02-01 06:49:27 
    L1   | 2017-02-01 09:35:39 
    L1   | 2017-02-01 09:51:41 
    //etc 
I want result in PostgreSQL like below, 

    time_range  | count 
    ---------------------------- 
    00:00-01:00 | 0 
    01:00-02:00 | 0 
    //etc 
    06:00-07:00 | 1 
    09:00-10:00 | 2

我想指望小時明智record.My確切概念是與count.I正在嘗試以下查詢，但不工作明智繪製圖表小時，PostgreSQL的範圍小時計數明智數據

select count(range) as ranges, 
    case 
    when to_char(updated_date,'HH:MI') >=00:00 and to_char(updated_date,'HH:MI')<=01:00 then '00:00-01:00' 
    when to_char(updated_date,'HH:MI') >=01:00 and to_char(updated_date,'HH:MI')<=02:00 then '01:00-02:00' 
    when to_char(updated_date,'HH:MI') >=02:00 and to_char(updated_date,'HH:MI')<=03:00 then '02:00-03:00' 
    when to_char(updated_date,'HH:MI') >=03:00 and to_char(updated_date,'HH:MI')<=04:00 then '03:00-04:00' 
    when to_char(updated_date,'HH:MI') >=04:00 and to_char(updated_date,'HH:MI')<=05:00 then '04:00-05:00' 
    when to_char(updated_date,'HH:MI') >=05:00 and to_char(updated_date,'HH:MI')<=06:00 then '05:00-06:00' 
    when to_char(updated_date,'HH:MI') >=06:00 and to_char(updated_date,'HH:MI')<=07:00 then '06:00-07:00' 
    when to_char(updated_date,'HH:MI') >=07:00 and to_char(updated_date,'HH:MI')<=08:00 then '07:00-08:00' 
    when to_char(updated_date,'HH:MI') >=08:00 and to_char(updated_date,'HH:MI')<=09:00 then '08:00-09:00' 
    when to_char(updated_date,'HH:MI') >=09:00 and to_char(updated_date,'HH:MI')<=10:00 then '09:00-10:00' 
    when to_char(updated_date,'HH:MI') >=10:00 and to_char(updated_date,'HH:MI')<=11:00 then '10:00-11:00' 
    when to_char(updated_date,'HH:MI') >=11:00 and to_char(updated_date,'HH:MI')<=12:00 then '11:00-12:00' 
    when to_char(updated_date,'HH:MI') >=12:00 and to_char(updated_date,'HH:MI')<=13:00 then '12:00-13:00' 
    //etc 
    else '' end AS range from 
    from my_table 
    where date(updated_date)=='2017-02-01'

來源

2017-02-24 bdevloper

你意識到點鐘位置出現在兩個範圍？你可能應該使用'和to_char（updated_date，'HH：MI'）<05：00' –

而......你的問題是什麼？你的查詢是否給出預期的結果？（OFC，修復了明顯的語法錯誤後，比如'> = 00：00'）如果不是，預期的結果是什麼？ – pozs

查詢沒有給出預期的結果。我想要明智地計數discount_code小時所以，我甘蔗用於繪圖圖 – bdevloper

什麼像這樣的事情，你可以用另一個日期函數格式化小時範圍。

DEMO

SELECT date_trunc('hour', "updated_date") as hour, count(*) 
FROM Table1 
GROUP BY date_trunc('hour', "updated_date") 
ORDER BY 1

輸出

編輯：更完整的版本，你需要計算範圍。

DEMO 2:

WITH parameter as (
    SELECT '2017-02-01'::date as d 
), hours (id, t) as (
    SELECT 1, '00:00'::time t UNION ALL 
    SELECT 2, '01:00'::time UNION ALL 
    SELECT 3, '02:00'::time UNION ALL 
    SELECT 4, '03:00'::time UNION ALL 
    SELECT 5, '04:00'::time UNION ALL 
    SELECT 6, '05:00'::time UNION ALL 
    SELECT 7, '06:00'::time UNION ALL 
    SELECT 8, '07:00'::time UNION ALL 
    SELECT 9, '08:00'::time UNION ALL 
    SELECT 10, '09:00'::time UNION ALL 
    SELECT 11, '10:00'::time UNION ALL 
    SELECT 12, '11:00'::time UNION ALL 
    SELECT 13, '12:00'::time UNION ALL 
    SELECT 14, '13:00'::time UNION ALL 
    SELECT 15, '14:00'::time UNION ALL 
    SELECT 16, '15:00'::time UNION ALL 
    SELECT 17, '16:00'::time UNION ALL 
    SELECT 18, '17:00'::time UNION ALL 
    SELECT 19, '18:00'::time UNION ALL 
    SELECT 20, '19:00'::time UNION ALL 
    SELECT 21, '20:00'::time UNION ALL 
    SELECT 22, '21:00'::time UNION ALL 
    SELECT 23, '22:00'::time UNION ALL 
    SELECT 24, '23:00'::time UNION ALL 
    SELECT 25, '24:00'::time 
), ranges as (
SELECT d + h1.t from_time, 
     d + h2.t to_time 
FROM parameter 
CROSS JOIN hours h1 
INNER JOIN hours h2 
     ON h1.id = h2.id - 1 
)  
SELECT from_time, to_time, count(t.updated_date) 
FROM ranges r 
LEFT JOIN Table1 t 
    ON t.updated_date >= r.from_time 
AND t.updated_date < r.to_time 
GROUP BY from_time, to_time 
ORDER BY from_time

輸出

來源

2017-02-24 15:00:25

胡安這是非常有用的答案。非常感謝你！ – bdevloper

嘗試：

select lpad(hr::text, 2, '0')||':00-'||lpad((hr+1)::text, 2, '0')||':00' as time_range, 
     count(discount_code) 
from (
     select * 
     from generate_series(0,23) hr 
) x 
left join my_table t on x.hr = extract(hour from t.UPDATED_DATE) 
group by lpad(hr::text, 2, '0')||':00-'||lpad((hr+1)::text, 2, '0')||':00'

來源

2017-02-24 15:15:31 krokodilko

謝謝krokodilko非常有用的答案 – bdevloper

使用generate_series()功能得到結果，你有沒有數據：

select to_char(h, 'HH24:MI-') || to_char(h + interval '1 hour', 'HH24:MI') time_range, 
      count(discount_code) 
from  generate_series(timestamp '2017-02-01', 
          timestamp '2017-02-02' - interval '1 hour', 
          interval '1 hour') h 
left join simple_table on updated_date >= h and updated_date < (h + interval '1 hour') 
group by h 
order by h;

http://rextester.com/EBBCL23865

來源

2017-02-24 15:20:35 pozs

pozs謝謝偉大 – bdevloper

PostgreSQL的範圍小時計數明智數據

回答

相關問題