I create a custom type dataset by spark.
public class KeyValuePair { String source; String target; int value; getter... setter... } ... List<KeyValuePair> list = generateList(); Dataset<KeyValuePair> dataset = spark.createDataset(list, Encoders.kryo(KeyValuePair.class)); Dataset<Row> agg = dataset.groupBy("source", "target").agg(avg("value"));
When the dataset calls the method groupBy, it will throw an exception. Exception in thread "main" org.apache.spark.sql.AnalysisException: Cannot resolve column name "source" among (value)
I wonder whether a custom type dataset can call the method groupBy. Is Row type dataset the only one can call this method without throw an exception?
How can I aggregate a custom type dataset?
https://stackoverflow.com/questions/66575894/how-does-a-custom-type-dataset-call-the-groupby-method March 11, 2021 at 11:08AM
没有评论:
发表评论