PySpark 使用 RDD 和 json.load 解析 Json

Question

{

  "city": "Tempe",
  "state": "AZ",
  ...
  "attributes": [
    "BikeParking: True",
    "BusinessAcceptsBitcoin: False",
    "BusinessAcceptsCreditCards: True",
    "BusinessParking: {'garage': False, 'street': False, 'validated': False, 'lot': True, 'valet': False}",
    "DogsAllowed: False",
    "RestaurantsPriceRange2: 2",
    "WheelchairAccessible: True"
  ],
  ...
}

你好，我正在使用 PySpark，我正在嘗試輸出一個 (state, BusinessAcceptsBitcoin) 元組，目前我正在做：

csr = (dataset
        .filter(lambda e:"city" in e and "BusinessAcceptsBitcoin" in e)
        .map(lambda e: (e["city"],e["BusinessAcceptsBitcoin"]))
        .collect()
        )

但是這個命令失敗了。 如何獲得“BusinessAcceptsBitcoin”和“城市”字段？

Answer 1

您可以使用 Dataframe 和 UDF 來解析“屬性”字符串。

從您提供的示例數據來看，“屬性”似乎不是正確的 JSON 或 Dict。

假設“屬性”只是一個字符串，這里是一個使用數據框和 Udf 的示例代碼。

from pyspark.sql import SparkSession
from pyspark.sql.functions import *
from pyspark.sql.types import *

spark = SparkSession \
            .builder \
            .appName("test") \
            .getOrCreate()

#sample data
data=[{

  "city": "Tempe",
  "state": "AZ",
  "attributes": [
    "BikeParking: True",
    "BusinessAcceptsBitcoin: False",
    "BusinessAcceptsCreditCards: True",
    "BusinessParking: {'garage': False, 'street': False, 'validated': False, 'lot': True, 'valet': False}",
    "DogsAllowed: False",
    "RestaurantsPriceRange2: 2",
    "WheelchairAccessible: True"
  ]
}]
df=spark.sparkContext.parallelize(data).toDF()

用戶定義的函數來解析字符串

def get_attribute(data,attribute):
    return [list_item for list_item in data if attribute in list_item][0]

注冊udf

udf_get_attribute=udf(get_attribute, StringType

數據框

df.withColumn("BusinessAcceptsBitcoin",udf_get_attribute("attributes",lit("BusinessAcceptsBitcoin"))).select("city","BusinessAcceptsBitcoin").show(truncate=False)

樣本輸出

+-----+-----------------------------+
|city |BusinessAcceptsBitcoin       |
+-----+-----------------------------+
|Tempe|BusinessAcceptsBitcoin: False|
+-----+-----------------------------+

例如，您也可以使用相同的 udf 來查詢任何其他字段

df.withColumn("DogsAllowed",udf_get_attribute("attributes",lit("DogsAllowed"))).select("city","DogsAllowed").show(truncate=False)