# 数据的分组聚合,找到每个用户最近的3次收藏beat(用window开窗函数) from pyspark.sql.window import Window import pyspark.sql.functions as F window_out = Window.partitionBy("user_id") \ .orderBy(F.desc("collect_time")) # user_feed.withColumn("rank", F.rank().over(window_out)).show(trunc...