[
https://issues.apache.org/jira/browse/SPARK-22505?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=16249702#comment-16249702
]
Ruslan Dautkhanov commented on SPARK-22505:
-------------------------------------------
In a way, you can think of this as of Pandas' infer_dtype() call:
https://pandas.pydata.org/pandas-docs/stable/generated/pandas.api.types.infer_dtype.html?highlight=infer#pandas.api.types.infer_dtype
One workaround for this missing Spark functionality is writing file back as
delimited, and then read it back in so we can use spark-csv schema inference.
But this would be super inefficient. Again, would be great to somehow engage
the same type inference as in spark-csv from an RDD of arbitrary tuples of
strings (or arrays).
> toDF() / createDataFrame() type inference doesn't work as expected
> ------------------------------------------------------------------
>
> Key: SPARK-22505
> URL: https://issues.apache.org/jira/browse/SPARK-22505
> Project: Spark
> Issue Type: Bug
> Components: PySpark, Spark Core
> Affects Versions: 2.2.0
> Reporter: Ruslan Dautkhanov
> Labels: csvparser, inference, pyspark, schema, spark-sql
>
> {code}
> df =
> sc.parallelize([('1','a'),('2','b'),('3','c')]).toDF(['should_be_int','should_be_str'])
> df.printSchema()
> {code}
> produces
> {noformat}
> root
> |-- should_be_int: string (nullable = true)
> |-- should_be_str: string (nullable = true)
> {noformat}
> Notice `should_be_int` has `string` datatype, according to documentation:
> https://spark.apache.org/docs/latest/sql-programming-guide.html#inferring-the-schema-using-reflection
> {quote}
> Spark SQL can convert an RDD of Row objects to a DataFrame, inferring the
> datatypes. Rows are constructed by passing a list of key/value pairs as
> kwargs to the Row class. The keys of this list define the column names of the
> table, *and the types are inferred by sampling the whole dataset*, similar to
> the inference that is performed on JSON files.
> {quote}
> Schema inference works as expected when reading delimited files like
> {code}
> spark.read.format('csv').option('inferSchema', True)...
> {code}
> but not when using toDF() / createDataFrame() API calls.
> Spark 2.2.
--
This message was sent by Atlassian JIRA
(v6.4.14#64029)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
