Python 2.7 Pyspark中的Py4JJavaError

我正在使用Python API开发Spark。下面是我的代码。当我执行wordCount.first行时。我收到ValueError:需要超过1个值才能解包。如能对上述错误有所了解，将不胜感激。谢谢

#create an RDD with textFile method
text_data_file=sc.textFile('/resources/yelp_labelled.txt')

#import the required library for word count operation
from operator import add
#Use filter to return RDD for words length greater than zero
wordCountFilter=text_data_file.filter(lambda x:len(x)>0)
#use flat map to split each line into words
wordFlatMap=wordCountFilter.flatMap(lambda x: x.split())
#map each key with value 1 using map function
wordMapper=wordFlatMap.flatMap(lambda x:(x,5))
#Use reducebykey function to reduce above mapped keys
#returns the key-value pairs by adding values for similar keys
wordCount=wordMapper.reduceByKey(add)
#view the first element
wordCount.first()

你的错误在这里：

wordMapper=wordFlatMap.flatMap(lambda x:(x,5))

应该是

wordMapper=wordFlatMap.map(lambda x:(x,5))

否则你只会发出

x

及

作为单独的值。Spark将尝试扩展x并失败，因为它的长度不等于2。否则，它将尝试解包5并失败

x

5