python, pyspark : get sum of a pyspark dataframe column values

4
votes

say I have a dataframe like this

name age city
abc   20  A
def   30  B

i want to add a summary row at the end of the dataframe, so result will be like

name age city
abc   20  A
def   30  B
All   50  All

So String 'All', I can easily put, but how to get sum(df['age']) ###column object is not iterable

data = spark.createDataFrame([("abc", 20, "A"), ("def", 30, "B")],["name", "age", "city"])
data.printSchema()
#root
 #|-- name: string (nullable = true)
 #|-- age: long (nullable = true)
 #|-- city: string (nullable = true)
res = data.union(spark.createDataFrame([('All',sum(data['age']),'All')], data.columns))  ## TypeError: Column is not iterable
#Even tried with data['age'].sum() and got error.   If i am using [('All',50,'All')], it is doing fine.

I usually work on Pandas dataframe and new to Spark. Might be my undestanding about spark dataframe is not that matured.

Please suggest, how to get the sum over a dataframe-column in pyspark. And if there is any better way to add/append a row to end of a dataframe. Thanks.

2
pythonpysparkpyspark-sql
Currently I am solving the above requirement by "sum_value = int(data.agg({'age':'sum'}).toPandas()['sum(age)'].sum())", i:e applying agg sum on column->converting to pandas df->on column/series applying sum function.. But i don't want to involve Pandas here. – Satya
thats the way I would choose: df.limit(20).agg(F.sum('count')).show() – pabloverd

2 Answers

15
votes

Spark SQL has a dedicated module for column functions pyspark.sql.functions.
So the way it works is:

from pyspark.sql import functions as F
data = spark.createDataFrame([("abc", 20, "A"), ("def", 30, "B")],["name", "age", "city"])

res = data.unionAll(
    data.select([
        F.lit('All').alias('name'), # create a cloumn named 'name' and filled with 'All'
        F.sum(data.age).alias('age'), # get the sum of 'age'
        F.lit('All').alias('city') # create a column named 'city' and filled with 'All'
    ]))
res.show()

Prints:

+----+---+----+
|name|age|city|
+----+---+----+
| abc| 20|   A|
| def| 30|   B|
| All| 50| All|
+----+---+----+
4
votes

A dataframe is immutable, you need to create a new one. To get the sum of your age, you can use this function: data.rdd.map(lambda x: float(x["age"])).reduce(lambda x, y: x+y)

The way you add a row is fine, but why would you do such a thing? Your dataframe will be hard to manipulate and you wont be able to use aggregations functions unless you drop the last line.