Question

我有一个表毒性，显示药物对几名患者的副作用，如下所示：

列中的值：恶心，呕吐和皮肤是指药物毒性。严重性有4个值：0,1,2,3

|-----|---------------|--------|-------|------| 
| ID  | Regime Course | Nausea | Vomit | Skin | 
| 1   | Regime A      | 1      | 2     | 1    | 
| 2   | Regime A      | 1      | 1     | 1    | 
| 3   | Regime A      | 2      | 3     | 2    | 
| 4   | Regime A      | 0      | 0     | 2    | 
| 5   | Regime B      | 3      | 0     | 3    | 
| 6   | Regime B      | 3      | 3     | 3    | 
| 7   | Regime B      | 1      | 3     | 1    | 
| 8   | Regime B      | 2      | 2     | 0    | 
| 9   | Regime A      | 3      | 2     | 0    | 
| 10  | Regime A      | 3      | 2     | 0    | 
| 11  | Regime A      | 0      | 1     | 1    | 
| 12  | Regime A      | 0      | 1     | 1    | 
| 13  | Regime B      | 1      | 1     | 2    | 
| 14  | Regime B      | 2      | 0     | 0    | 
| 15  | Regime B      | 1      | 0     | 2    | 
| 16  | Regime B      | 0      | 0     | 0    | 
|-----|---------------|--------|-------|------|

我应该使用什么查询，以便我可以获得两种方案的所有副作用的百分比（例如，恶心= 1，恶心= 2，皮肤= 0，皮肤= 2）？

结果示例：

Regime A |  Nausea:0  | 0.375
Regime A |  Nausea:1  | 0.25
Regime A |  Nausea:2  | 0.125
Regime A |  Nausea:3  | 0.25
Regime B |  Nausea:0  | 0.125
Regime B |  Nausea:1  | 0.375
Regime B |  Nausea:2  | 0.25
Regime B |  Nausea:3  | 0.25
...
...

谢谢！

Answer 1

以下是获得恶心比率的解决方案：

select
  regime,
  nausea,
  regime_nausea_count / (1.0*regime_count) as nausea_ratio
from 
  (
    select
      regime,
      nausea,
      count(*) as regime_nausea_count
    from
      Toxicity
    group by regime, nausea
   ),
 (
    select
      regime,
      count(*) as regime_count
    from
      Toxicity
    group by regime
  ) using (regime);

显然，您可以用其他列替换恶心来获取相应的统计信息。

如何创建查询以查找数据百分比（基于每个字段的值）？

1 个答案: