我有下表：

+---------+---------+--------+------+
| Whisper | Client  | Tenant | Ring |
+---------+---------+--------+------+
| W1      | iOS     | T3     | R1   |
+---------+---------+--------+------+
| W2      | iOS     | T2     | R1   |
+---------+---------+--------+------+
| W1      | Android | T3     | R1   |
+---------+---------+--------+------+
| W2      | Android | T2     | R1   |
+---------+---------+--------+------+
| W3      | Android | T4     | R2   |
+---------+---------+--------+------+

我想计算清楚客户栏有变化时的不同耳语，目前我使用的是：

WhipserCounts =
SELECT COUNT(DISTINCT Whipser) AS Whispers,
                Client,
                Tenant,
                Ring
FROM InputData
GROUP BY Client, Tenant, Ring;

但是我看到w1和w2对每个客户机都计算了两次，但是无论客户机是什么，我都需要将它们视为一个，所以第一次如果我得到ios，我需要将客户机设置为第一个值(知道我同时需要输出表中的客户机）。
我得到的是：

+----------+---------+--------+------+
| Whispers | Client  | Tenant | Ring |
+----------+---------+--------+------+
| 1        | iOS     | T3     | R1   |
+----------+---------+--------+------+
| 1        | iOS     | T2     | R1   |
+----------+---------+--------+------+
| 1        | Android | T3     | R1   |
+----------+---------+--------+------+
| 1        | Android | T2     | R1   |
+----------+---------+--------+------+
| 1        | Android | T4     | R2   |
+----------+---------+--------+------+

我需要的是：

+----------+---------+--------+------+
| Whispers | Client  | Tenant | Ring |
+----------+---------+--------+------+
| 1        | iOS     | T3     | R1   |
+----------+---------+--------+------+
| 1        | iOS     | T2     | R1   |
+----------+---------+--------+------+
| 1        | Android | T4     | R2   |
+----------+---------+--------+------+

你能帮个忙吗？