Question

我的参考是Transpose rows into columns in BigQuery (Pivot implementation)

我尝试通过用string_agg替换group_concat_unquoted来在标准的大查询中实施建议，但是我没有成功使用正确的语法和结构。

WITH
input AS (
SELECT
    user_pseudo_id AS user_id,
    event_timestamp,
    p.key AS key,
    p.value.string_value AS value
  FROM
    `myproject.analytics_xxxx.events_*`,
    UNNEST(event_params) AS p
  WHERE
    AND event_name = 'event_type_to_pivot' )
SELECT
  'select user_id, ' + STRING_AGG(
  'MAX(IF(key = "'+key+'", value, NULL)) as ['+key+']'
  ) + 'FROM input group by user_id order by user_id'
FROM (
  SELECT
    key
  FROM
    input
  GROUP BY
    key
  ORDER BY
    key )

我总是遇到错误：对于参数类型：STRING，STRING，运算符+没有匹配的签名。支持的签名：INT64 + INT64； FLOAT64 + FLOAT64; NUMERIC + NUMERIC在[19：3]行：

STRING_AGG(
      'MAX(IF(key = "'+key+'", value, NULL)) as ['+key+']'
      )

经过大量尝试，我仍然不了解如何正确使用string_agg将行转置为列。

感谢您的指导，

Answer 1

在#standardSQL中，使用private function getDataHelper(Request $request) { $message = new Message([ 'title' => $request->get('title'), 'description' => $request->get('description'), 'branch_id' => Auth::user()->branch_id, 'user_id' => Auth::user()->id ]); $title = $request->get('title'); $description = $request->get('description'); if($title === 'products' || $description === 'products') { return redirect('getproducts'); } if($title === 'customer' || $description === 'customer') { return redirect('getcustomer'); } if($title === 'shop' || $description === 'shop') { return redirect('getshop'); } // otherwise redirect home return redirect('/'); }代替CONCAT()来组成字符串：

在标准BigQuery（数据透视实现）中将行转置为列

1 个答案: