PySpark Dataframe.groupBy MapType Column

Question

I have a dataframe with a MapType column where the key is id and the value is another StructType with two numbers, a counter and an income.

It looks like this:

+--------------------------------------+
| myMapColumn                          |
+--------------------------------------+
| Map(1 -> [1, 4.0], 2 -> [1, 1.5])    |
| Map()                                |
| Map(1 -> [3, 5.5])                   |
| Map(1 -> [4, 0.1], 2 -> [6, 101.56]) |
+--------------------------------------+

Now I need to sum these two values ​​for the identifier, and the result will be:

+----------------------+
| id | count | revenue |
+----------------------+
| 1  | 8     | 9.6     |
| 2  | 7     | 103.06  |
+----------------------+

I really don’t know how to do this, and could not find the documentation for this special case. I tried to use Dataframe.groupBy but could not get it to work :(

Any ideas?

I am using Spark 1.5.2 with Python 2.6.6

Answer 1

Assuming the circuit is equivalent to this:

root
 |-- myMapColumn: map (nullable = true)
 |    |-- key: integer
 |    |-- value: struct (valueContainsNull = true)
 |    |    |-- _1: integer (nullable = false)
 |    |    |-- _2: double (nullable = false)

, , explode :

from pyspark.sql.functions import col, explode, sum as sum_

(df
  .select(explode(col("myMapColumn")))
  .groupBy(col("key").alias("id"))
  .agg(sum_("value._1").alias("count"), sum_("value._2").alias("revenue")))