Pyspark:通过检查值是否存在来聚合数据(不是计数或总和)

cqc*_*991 5 sql pyspark

我有一个这样的数据集,

test = spark.createDataFrame([
    (0, 1, 5, "2018-06-03", "Region A"),
    (1, 1, 2, "2018-06-04", "Region B"),
    (2, 2, 1, "2018-06-03", "Region B"),
    (4, 1, 1, "2018-06-05", "Region C"),
    (5, 3, 2, "2018-06-03", "Region D"),
    (6, 1, 2, "2018-06-03", "Region A"),
    (7, 4, 4, "2018-06-03", "Region A"),
    (8, 4, 4, "2018-06-03", "Region B"),
    (9, 5, 4, "2018-06-03", "Region A"),
    (10, 5, 4, "2018-06-03", "Region B"),
])\
  .toDF("orderid", "customerid", "price", "transactiondate", "location")
test.show()
Run Code Online (Sandbox Code Playgroud)

我可以像这样汇总每个客户在每个区域的订单:

temp_result = test.groupBy("customerid").pivot("location").agg(count("orderid")).na.fill(0)
temp_result.show()
Run Code Online (Sandbox Code Playgroud)

在此输入图像描述

现在,我想通过确定值是否存在(即 0 或 1)来简单地聚合数据,而不是 or ,如下sum所示count

在此输入图像描述


我可以通过以下方式获得上述结果

for field in temp_result.schema.fields:
    if str(field.name) not in ['customerid', "overall_count", "overall_amount"]:
        name = str(field.name)
        temp_result = temp_result.withColumn(name, \
                                             when(col(name) >= 1, 1).otherwise(0))
Run Code Online (Sandbox Code Playgroud)

但有没有更简单的方法来获取它?

pan*_*sen 4

您基本上已经完成了 - 只需进行一点调整即可获得您想要的结果。在聚合中,添加计数比较并将布尔值转换为整数(如果有必要):

temp_result = test.groupBy("customerid")\
                  .pivot("location")\
                  .agg((count("orderid")>0).cast("integer"))\
                  .na.fill(0)

temp_result.show()
Run Code Online (Sandbox Code Playgroud)

结果为:

+----------+--------+--------+--------+--------+
|customerid|Region A|Region B|Region C|Region D|
+----------+--------+--------+--------+--------+
|         5|       1|       1|       0|       0|
|         1|       1|       1|       1|       0|
|         3|       0|       0|       0|       1|
|         2|       0|       1|       0|       0|
|         4|       1|       1|       0|       0|
+----------+--------+--------+--------+--------+
Run Code Online (Sandbox Code Playgroud)

如果出现 Spark 错误,您可以使用此解决方案,它通过附加步骤进行计数比较:

temp_result = test.groupBy("customerId", "location")\
                  .agg(count("orderid").alias("count"))\
                  .withColumn("count", (col("count")>0).cast("integer"))\
                  .groupby("customerId")\
                  .pivot("location")\
                  .agg(sum("count")).na.fill(0)

temp_result.show()
Run Code Online (Sandbox Code Playgroud)