Spark / Parquet分区是否保持顺序?

时间:2019-03-07 23:12:36

标签: apache-spark pyspark parquet

如果我对数据集进行分区,当我读回它时,它的顺序是否正确?例如,考虑以下pyspark代码:

# read a csv
df = sql_context.read.csv(input_filename)

# add a hash column
hash_udf = udf(lambda customer_id: hash(customer_id) % 4, IntegerType())
df = df.withColumn('hash', hash_udf(df['customer_id']))

# write out to parquet
df.write.parquet(output_path, partitionBy=['hash'])

# read back the file
df2 = sql_context.read.parquet(output_path)

我在customer_id存储桶上进行分区。当我读回整个数据集时,是否保证分区可以按照原始插入顺序重新合并在一起?

现在,我不太确定,所以我要添加一个序列列:

df = df.withColumn('seq', monotonically_increasing_id())

但是,我不知道这是否多余。

1 个答案:

答案 0 :(得分:2)

不,不能保证。尝试使用很小的数据集:

df = spark.createDataFrame([(1,'a'),(2,'b'),(3,'c'),(4,'d')],['customer_id', 'name'])

# add a hash column
hash_udf = udf(lambda customer_id: hash(customer_id) % 4, IntegerType())
df = df.withColumn('hash', hash_udf(df['customer_id']))

# write out to parquet
df.write.parquet("test", partitionBy=['hash'], mode="overwrite")

# read back the file
df2 = spark.read.parquet("test")
df.show()

+-----------+----+----+
|customer_id|name|hash|
+-----------+----+----+
|          1|   a|   1|
|          2|   b|   2|
|          3|   c|   3|
|          4|   d|   0|
+-----------+----+----+
df2.show()

+-----------+----+----+
|customer_id|name|hash|
+-----------+----+----+
|          2|   b|   2|
|          1|   a|   1|
|          4|   d|   0|
|          3|   c|   3|
+-----------+----+----+