I think it works in Python
```
>>> df = sqlContext.createDataFrame([(1, {'a': 1})])
>>> df.printSchema()
root
|-- _1: long (nullable = true)
|-- _2: map (nullable = true)
| |-- key: string
| |-- value: long (valueContainsNull = true)
>>> df.select(df._2.getField('a')).show()
+-----+
|_2[a]|
+-----+
| 1|
+-----+
>>> df.select(df._2['a']).show()
+-----+
|_2[a]|
+-----+
| 1|
+-----+
```
On Mon, Jun 8, 2015 at 6:00 AM, mrm <[email protected]> wrote:
> Hi,
>
> Is it possible to query a data structure that is a dictionary within a
> dictionary?
>
> I have a parquet file with a a structure:
> test
> |____key1: {key_string: val_int}
> |____key2: {key_string: val_int}
>
> if I try to do:
> parquetFile.test
> --> Column<test>
>
> parquetFile.test.key2
> --> AttributeError: 'Column' object has no attribute 'key2'
>
> Similarly, if I try to do a SQL query, it throws this error:
>
> org.apache.spark.sql.AnalysisException: GetField is not valid on fields of
> type MapType(StringType,MapType(StringType,IntegerType,true),true);
>
> Is this at all possible with the Python API in Spark SQL?
>
> Thanks,
> Maria
>
>
>
> --
> View this message in context:
> http://apache-spark-user-list.1001560.n3.nabble.com/SparkSQL-nested-dictionaries-tp23207.html
> Sent from the Apache Spark User List mailing list archive at Nabble.com.
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [email protected]
> For additional commands, e-mail: [email protected]
>
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
