如何从pyspark.sql.function中提取值

Question

我正在使用一些pyspark.sql.functions:

print(ratings.select(mean('rating')).take(1))
print(ratings.select(stddev('rating')).take(1))

输出为:

[Row(avg(rating)=3.581564453029317)]
[Row(stddev_samp(rating,0,0)=1.1171018453732544)]

如何提取值，以便我可以将其分配给变量，例如

mean_ratings = ratings.select(mean('rating'))

Answer 1

Take返回一个行列表。索引到列表中以获得第一行，然后拉出您正在查找的字段:

mean_ratings = ratings.select(mean('rating')).take(1)[0]['avg(rating)']