You are viewing a plain text version of this content. The canonical link for it is here.
Posted to issues@spark.apache.org by "Xiao Li (JIRA)" <ji...@apache.org> on 2018/10/29 00:58:00 UTC
[jira] [Resolved] (SPARK-25816) Functions does not resolve Columns
correctly
[ https://issues.apache.org/jira/browse/SPARK-25816?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel ]
Xiao Li resolved SPARK-25816.
-----------------------------
Resolution: Fixed
Fix Version/s: 2.4.0
2.3.3
> Functions does not resolve Columns correctly
> --------------------------------------------
>
> Key: SPARK-25816
> URL: https://issues.apache.org/jira/browse/SPARK-25816
> Project: Spark
> Issue Type: Bug
> Components: SQL
> Affects Versions: 2.3.0, 2.3.1, 2.3.2, 2.4.0
> Reporter: Brian Zhang
> Assignee: Peter Toth
> Priority: Critical
> Fix For: 2.3.3, 2.4.0
>
> Attachments: final_allDatatypes_Spark.avro, source.snappy.parquet
>
>
> When there is a duplicate column name in the current Dataframe and orginal Dataframe where current df is selected from, Spark in 2.3.0 and 2.3.1 does not resolve the column correctly when using it in the expression, hence causing casting issue. The same code is working in Spark 2.2.1
> Please see below code to reproduce the issue
> import org.apache.spark._
> import org.apache.spark.rdd._
> import org.apache.spark.storage.StorageLevel._
> import org.apache.spark.sql._
> import org.apache.spark.sql.DataFrame
> import org.apache.spark.sql.types._
> import org.apache.spark.sql.functions._
> import org.apache.spark.sql.catalyst.expressions._
> import org.apache.spark.sql.Column
> val v0 = spark.read.parquet("/data/home/bzinfa/bz/source.snappy.parquet")
> val v00 = v0.toDF(v0.schema.fields.indices.view.map("" + _):_*)
> val v5 = v00.select($"13".as("0"),$"14".as("1"),$"15".as("2"))
> val v5_2 = $"2"
> v5.where(lit(500).<(v5_2(new Column(new MapKeys(v5_2.expr))(lit(0)))))
> //v00's 3rdcolumn is binary and 16th is map<string, double>
> Error:
> org.apache.spark.sql.AnalysisException: cannot resolve 'map_keys(`2`)' due to data type mismatch: argument 1 requires map type, however, '`2`' is of binary type.;
>
> 'Project [0#1591, 1#1592, 2#1593] +- 'Filter (500 < {color:#FF0000}2#1593{color}[map_keys({color:#FF0000}2#1561{color})[0]]) +- Project [13#1572 AS 0#1591, 14#1573 AS 1#1592, 15#1574 AS 2#1593, 2#1561] +- Project [c_bytes#1527 AS 0#1559, c_union#1528 AS 1#1560, c_fixed#1529 AS 2#1561, c_boolean#1530 AS 3#1562, c_float#1531 AS 4#1563, c_double#1532 AS 5#1564, c_int#1533 AS 6#1565, c_long#1534L AS 7#1566L, c_string#1535 AS 8#1567, c_decimal_18_2#1536 AS 9#1568, c_decimal_28_2#1537 AS 10#1569, c_decimal_38_2#1538 AS 11#1570, c_date#1539 AS 12#1571, simple_struct#1540 AS 13#1572, simple_array#1541 AS 14#1573, simple_map#1542 AS 15#1574] +- Relation[c_bytes#1527,c_union#1528,c_fixed#1529,c_boolean#1530,c_float#1531,c_double#1532,c_int#1533,c_long#1534L,c_string#1535,c_decimal_18_2#1536,c_decimal_28_2#1537,c_decimal_38_2#1538,c_date#1539,simple_struct#1540,simple_array#1541,simple_map#1542] parquet
--
This message was sent by Atlassian JIRA
(v7.6.3#76005)
---------------------------------------------------------------------
To unsubscribe, e-mail: issues-unsubscribe@spark.apache.org
For additional commands, e-mail: issues-help@spark.apache.org