Serialize a Spark DataFrame into Apache Avro format.
Notice this functionality requires the Spark connection sc to be instantiated with either
an explicitly specified Spark version (i.e.,
spark_connect(..., version = <version>, packages = c("avro", <other package(s)>), ...))
or a specific version of Spark avro package to use (e.g.,
spark_connect(..., packages = c("org.apache.spark:spark-avro_2.12:3.0.0", <other package(s)>), ...)).
spark_write_avro(
x,
path,
avro_schema = NULL,
record_name = "topLevelRecord",
record_namespace = "",
compression = "snappy",
partition_by = NULL
)A Spark DataFrame or dplyr operation
The path to the file. Needs to be accessible from the cluster. Supports the "hdfs://", "s3a://" and "file://" protocols.
Optional Avro schema in JSON format
Optional top level record name in write result (default: "topLevelRecord")
Record namespace in write result (default: "")
Compression codec to use (default: "snappy")
A character vector. Partitions the output by the given columns on the file system.
Other Spark serialization routines:
collect_from_rds(),
spark_insert_table(),
spark_load_table(),
spark_read_avro(),
spark_read_binary(),
spark_read_csv(),
spark_read_delta(),
spark_read_image(),
spark_read_jdbc(),
spark_read_json(),
spark_read_libsvm(),
spark_read_orc(),
spark_read_parquet(),
spark_read_source(),
spark_read_table(),
spark_read_text(),
spark_read(),
spark_save_table(),
spark_write_csv(),
spark_write_delta(),
spark_write_jdbc(),
spark_write_json(),
spark_write_orc(),
spark_write_parquet(),
spark_write_source(),
spark_write_table(),
spark_write_text()