sql orderby limit 1 得到的结果有时候会不一样；
原因是基于非唯一字段的排序 结果集是不确定的，需要再增加一个唯一字段一起来做排序这样就会保证结果一致了；
第一条： select * from 表名 limit 1;（默认是升序排序，取第一条）
最后一条：select * from表名 order by 表_id desc limit 1（降序排序取第一条）
--找到与概率0.5最接近的方法1（得到的概率与方法2一样，只是排名不一样，看情况取值）
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE big_hy_name ='nan';
--5982 0.4746192893401015
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob < 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob DESC,sub_rank DESC LIMIT 1;
--5983 0.5041050903119869
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob >= 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob ASC,sub_rank ASC LIMIT 1;
--找到与概率0.5最接近的方法2
--5976 0.4746192893401015
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob < 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob DESC,sub_rank ASC LIMIT 1;
--5984 0.5041050903119869
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob >= 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob ASC,sub_rank DESC LIMIT 1;
--找概率的真实的上下界
--5978 0.4746192893401015
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob < 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob DESC;
--5983 0.5041050903119869
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob >= 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob ASC;
if CollectUtils.is_empty(row_left): # 左值未找到，取右值
return row_right[select_column]
elif CollectUtils.is_empty(row_right): # 右值未找到，取左值，因为是一个已知表，一定会存在一边有值；
return row_left[select_column]
else:
# 绝对差值
abs_left = abs(row_left['shixin_prob'] - prob)
abs_right = abs(row_right['shixin_prob'] - prob)
selected_row = row_left if abs_left < abs_right else row_right ##
#注：词句为了看它的覆盖率，可以拆分成if...else...:
if abs_left < abs_right:
selected_row = row_left
else:
selected_row = row_right
return selected_row[select_column] #返回得到的sub_rank值；SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob < 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob DESC LIMIT 1; 5996
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob >= 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob ASC LIMIT 1; 5983
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob < 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob,sub_rank DESC LIMIT 1; 468
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE shixin_prob >= 0.5 AND big_hy_name ='nan' ORDER BY shixin_prob,sub_rank ASC LIMIT 1; 5983
SELECT sub_rank,shixin_prob FROM shixin_offline_predict_results WHERE big_hy_name ='nan' #
已知sub_rank是升序，与第二个加了sub_rank中的顺序ASC一致，所以才不会错？？？、DBA在和开发沟通/培训过程中要注意这一点 基于非唯一字段的排序 结果集是不确定的。如果业务逻辑对分页或者order by结果集有比较高的严格要求 ，请记得利用唯一键排序。
select sub_rank from t_features order by prob limit 1
assert检测sub_rank的时候 线上是4834，而先上是4833，
这种语句会导致测试线上线下通不过，原因是基于非唯一字段的排序 结果集是不确定的，需要再增加一个唯一字段一起来做排序这样就会保证结果一致了；
Order by和Limit混合使用引起的问题
如果在order by语句中返回的结果集有很多行，那么非排序的列的返回结果是不确定的，即随机的，所以如果配合limit的话每次返回的结果集的顺序是不固定的，比如下面这个例子
mysql> SELECT * FROM ratings ORDER BY category;
+----+----------+--------+
| id | category | rating |
+----+----------+--------+
| 1 | 1 | 4.5 |
| 5 | 1 | 3.2 |
| 3 | 2 | 3.7 |
| 4 | 2 | 3.5 |
| 6 | 2 | 3.5 |
| 2 | 3 | 5.0 |
| 7 | 3 | 2.7 |
+----+----------+--------+
使用了limit以后，可发现id列和rating列和之前的结果集顺序有出入：
mysql> SELECT * FROM ratings ORDER BY category LIMIT 5;
+----+----------+--------+
| id | category | rating |
+----+----------+--------+
| 1 | 1 | 4.5 |
| 5 | 1 | 3.2 |
| 4 | 2 | 3.5 |
| 3 | 2 | 3.7 |
| 6 | 2 | 3.5 |
+----+----------+--------+
如果你有必要保证每次有相同的结果集，则需要order by你需要的那几列了：
mysql> SELECT * FROM ratings ORDER BY category, id;
+----+----------+--------+
| id | category | rating |
+----+----------+--------+
| 1 | 1 | 4.5 |
| 5 | 1 | 3.2 |
| 3 | 2 | 3.7 |
| 4 | 2 | 3.5 |
| 6 | 2 | 3.5 |
| 2 | 3 | 5.0 |
| 7 | 3 | 2.7 |
+----+----------+--------+
mysql> SELECT * FROM ratings ORDER BY category, id LIMIT 5;
+----+----------+--------+
| id | category | rating |
+----+----------+--------+
| 1 | 1 | 4.5 |
| 5 | 1 | 3.2 |
| 3 | 2 | 3.7 |
| 4 | 2 | 3.5 |
| 6 | 2 | 3.5 |
+----+----------+--------+