消息“No matching signature for operator = for argument types：字符串，INT64.支持的签名：ANY = ANY”表示station_id和start_station_id列具有不同的数据类型，一个是字符串，另一个是整数。
若要修复此问题，请将其中一列强制转换为另一列的数据类型。你需要弄清楚哪个是整数，哪个是字符串。例如

SELECT data_type
FROM `bigquery-public-data.new_york_citibike.INFORMATION_SCHEMA.COLUMNS`
WHERE table_name = 'citibike_stations' AND column_name = 'station_id'

我建议将整数列转换为字符串（所有整数都可以是字符串，但不能反过来）

SELECT
       station_id
     , name
     , number_of_rides AS number_of_rides_starting_at_station
FROM (
  SELECT start_station_id, COUNT(*) number_of_rides
  FROM bigquery-public-data.new_york_citibike.citibike_trips
  GROUP BY start_station_id
) subquery
INNER JOIN bigquery-public-data.new_york_citibike.citibike_stations
    ON CAST(station_id AS STRING) = subquery.start_station_id
ORDER BY number_of_rides DESC

当然，这是一个猜测，它可能是相反的方式，即：

INNER JOIN bigquery-public-data.new_york_citibike.citibike_stations
    ON station_id = CAST(subquery.start_station_id AS STRING)

nb：可以使用safe_cast（）进行转换，但不需要将integer转换为string