从另一个DataFrame添加一列 [英] Add a column from another DataFrame

问题描述

在Scala Spark中，我可以轻松地将列添加到现有的Dataframe文字中

In Scala Spark, I can easily add a column to an existing Dataframe writing

val newDf = df.withColumn("date_min", anotherDf("date_min"))

在PySpark中这样做会导致AnalysisException.

Doing so in PySpark results in an AnalysisException.

这是我在做什么:

minDf.show(5)
maxDf.show(5)
+--------------------+
|            date_min|
+--------------------+
|2016-11-01 10:50:...|
|2016-11-01 11:46:...|
|2016-11-01 19:23:...|
|2016-11-01 17:01:...|
|2016-11-01 09:00:...|
+--------------------+
only showing top 5 rows

+--------------------+
|            date_max|
+--------------------+
|2016-11-01 10:50:...|
|2016-11-01 11:46:...|
|2016-11-01 19:23:...|
|2016-11-01 17:01:...|
|2016-11-01 09:00:...|
+--------------------+
only showing top 5 rows

然后，导致错误的结果:

And then, what results in an error :

newDf = minDf.withColumn("date_max", maxDf["date_max"])

AnalysisExceptionTraceback (most recent call last)
<ipython-input-13-7e19c841fa51> in <module>()
      2 maxDf.show(5)
      3 
----> 4 newDf = minDf.withColumn("date_max", maxDf["date_max"])

/opt/spark-2.1.0-bin-hadoop2.7/python/pyspark/sql/dataframe.pyc in withColumn(self, colName, col)
   1491         """
   1492         assert isinstance(col, Column), "col should be Column"
-> 1493         return DataFrame(self._jdf.withColumn(colName, col._jc), self.sql_ctx)
   1494 
   1495     @ignore_unicode_prefix

/opt/spark-2.1.0-bin-hadoop2.7/python/lib/py4j-0.10.4-src.zip/py4j/java_gateway.py in __call__(self, *args)
   1131         answer = self.gateway_client.send_command(command)
   1132         return_value = get_return_value(
-> 1133             answer, self.gateway_client, self.target_id, self.name)
   1134 
   1135         for temp_arg in temp_args:

/opt/spark-2.1.0-bin-hadoop2.7/python/pyspark/sql/utils.pyc in deco(*a, **kw)
     67                                              e.java_exception.getStackTrace()))
     68             if s.startswith('org.apache.spark.sql.AnalysisException: '):
---> 69                 raise AnalysisException(s.split(': ', 1)[1], stackTrace)
     70             if s.startswith('org.apache.spark.sql.catalyst.analysis'):
     71                 raise AnalysisException(s.split(': ', 1)[1], stackTrace)

AnalysisException: u'resolved attribute(s) date_max#67 missing from date_min#66 in operator !Project [date_min#66, date_max#67 AS date_max#106];;\n!Project [date_min#66, date_max#67 AS date_max#106]\n+- Project [date_min#66]\n   +- Project [cast((cast(date_min#6L as double) / cast(1000 as double)) as timestamp) AS date_min#66, cast((cast(date_max#7L as double) / cast(1000 as double)) as timestamp) AS date_max#67]\n      +- SubqueryAlias df, `df`\n         +- LogicalRDD [idvisiteur#5, date_min#6L, date_max#7L, sales_sum#8, sales_count#9L]\n'

推荐答案

希望这会有所帮助！

from pyspark.sql.functions import monotonically_increasing_id, row_number
from pyspark.sql.window import Window

minDf = sc.parallelize([['2016-11-01 10:50:00'],['2016-11-01 11:46:00']]).toDF(["date_min"])
maxDf = sc.parallelize([['2016-11-01 10:50:00'],['2016-11-01 11:46:00']]).toDF(["date_max"])

# since there is no common column between these two dataframes add row_index so that it can be joined
minDf=minDf.withColumn('row_index', row_number().over(Window.orderBy(monotonically_increasing_id())))
maxDf=maxDf.withColumn('row_index', row_number().over(Window.orderBy(monotonically_increasing_id())))

minDf = minDf.join(maxDf, on=["row_index"]).drop("row_index")
minDf.show()

输出为:

+-------------------+-------------------+
|           date_min|           date_max|
+-------------------+-------------------+
|2016-11-01 10:50:00|2016-11-01 10:50:00|
|2016-11-01 11:46:00|2016-11-01 11:46:00|
+-------------------+-------------------+

这篇关于从另一个DataFrame添加一列的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！