创建将参数显示为列的查询 - BigQuery答案

【问题标题】：Creating a query that shows parameters as columns - BigQuery创建将参数显示为列的查询 - BigQuery
【发布时间】：2019-01-28 15:39:17
【问题描述】：

我正在使用 Firebase 和 BigQuery。我试图做的是创建查询，显示平均数量的“订单”和每个“级别”的“软货币”，但我不知道该怎么做。

这是我数据库中示例事件的简化 JSON 文件。每次玩家达到下一个级别时都会发送此类事件。

[
  {
    "event_name": "LevelUp",
    "event_params": [
      {
        "key": "Level",
        "value": {
          "string_value": null,
          "int_value": "12"
        }
      },
      {
        "key": "SoftCurrrency",
        "value": {
          "string_value": null,
          "int_value": "623"
        }
      },
      {
        "key": "Orders",
        "value": {
          "string_value": null,
          "int_value": "24"
        }
      }
    ]
  }
]

【问题讨论】：

您必须获取所有数据并通过代码进行查询/过滤

标签： sql firebase google-bigquery

【解决方案1】：

我就是这样做的：

SELECT ANY_VALUE(Level) as Lvl, FLOOR(AVG(SoftCurrency)) as average_Coins, FLOOR(AVG(Orders)) as average_Orders
FROM  (
  SELECT MAX(if(param.key = "SoftCurrency", param.value.int_value, NULL)) AS SoftCurrency,
    MAX(if(param.key = "Level", param.value.int_value, NULL)) AS Level,
    MAX(if(param.key = "Orders", param.value.int_value, NULL)) AS Orders
  FROM (
    SELECT user_pseudo_id, event_timestamp, param
    FROM `analytics`,
    UNNEST(event_params) AS param
    WHERE event_name = "LevelUp"
    AND (param.key = "SoftCurrency" OR param.key = "Level" OR param.key = "Orders")
  )
  GROUP BY user_pseudo_id, event_timestamp
) 
GROUP BY Level
ORDER BY Lvl

【讨论】：