MySQL Workbench 6.3。我有以下数据集。
id date value
36 12/1/2015 3174
36 11/1/2015 3143
36 10/1/2015 3112
36 9/1/2015 3082
36 8/1/2015 3052
36 7/1/2015 3021
36 6/1/2015 2990
64 11/1/2015 3105
64 10/1/2015 3074
64 8/1/2015 3014
64 7/1/2015 2983
64 6/1/2015 2952
65 12/1/2015 3414
72 10/1/2015 3352
72 9/1/2015 3322
72 8/1/2015 3292
72 7/1/2015 3261
72 6/1/2015 3230
72 5/1/2015 3198
72 4/1/2015 3169
72 3/1/2015 3139
72 2/1/2015 3107
72 1/1/2015 3079
我想要的是按ID分组并获得最后3个月的值。 (如果原始数据中没有记录,请保留所有日期和值。) 下表是我的手动输出,以显示我想要的内容。非常感谢你。
id current_month value1 1_month_before_current value2 2_month_before_current value3 3_month_before_current value3
36 12/1/2015 3174 11/1/2015 3143 10/1/2015 3112 9/1/2015 3082
64 null null 11/1/2015 3105 10/1/2015 3074 null null
72 null null null null 10/1/2015 3352 9/1/2015 3322
答案 0 :(得分:0)
只需使用条件聚合:
select id,
max(case when date = '2015-12-01' then date end) as current_month,
max(case when date = '2015-12-01' then value end) as current_value,
max(case when date = '2015-11-01' then date end) as prev_month,
max(case when date = '2015-11-01' then value end) as prev_value,
max(case when date = '2015-10-01' then date end) as prev2_month,
max(case when date = '2015-10-01' then value end) as prev2_value,
from t
group by id;
如果您不想输入日期:
select id,
max(case when date = curmon then date end) as current_month,
max(case when date = curmon then value end) as current_value,
max(case when date = curmon - interval 1 month then date end) as prev_month,
max(case when date = curmon - interval 1 month then value end) as prev_value,
max(case when date = curmon - interval 2 month then date end) as prev2_month,
max(case when date = curmon - interval 2 month then value end) as prev2_value,
from t cross join
(select date('2015-12-01') as curmon) params
group by id;
此外,如果日期不是全月的第一天,您可以在case
表达式中使用范围逻辑。