pyspark throws TypeError: textFile() missing 1 required positional argument: 'name'

Question

I googled this problem, yet no direct answer related to spark-2.2.0-bin-hadoop2.7. I am trying to read a text file from local directory, but I always get TypeError that name argument is missing. This is the code in jupyter notebook with Python3:

from pyspark import SparkContext as sc
data = sc.textFile("/home/bigdata/test.txt")

When I run the cell, I get this error:

TypeError                                 Traceback (most recent call last)
  <ipython-input-7-2a326e5b8f8c> in <module>()
  1 from pyspark import SparkContext as sc
  ----> 2 data = sc.textFile("/home/bigdata/test.txt")
  TypeError: textFile() missing 1 required positional argument: 'name'

感谢您的帮助。

Answer 1

您正在调用textFile()实例方法

def textFile(self, name, minPartitions=None, use_unicode=True):

就像它是一个静态方法一样，导致将"/home/bigdata/test.txt"字符串用作self值，name而未指定参数，从而导致错误。

创建SparkContext该类的实例：

from pyspark import SparkConf
from pyspark.context import SparkContext

sc = SparkContext.getOrCreate(SparkConf().setMaster("local[*]"))
data = sc.textFile("/home/bigdata/test.txt")

Answer 2

from pyspark import SparkConf
from pyspark.context import SparkContext
sc = SparkContext.getOrCreate(SparkConf())
data = sc.textFile("my_file.txt")

显示一些内容

['这是文本文件，sc 工作正常']