I am running Spark 2.1 with Scala. I am trying to convert and array of vectors into a DenseVector.
Here is my dataframe:
scala> df_transformed.printSchema()
root
|-- id: long (nullable = true)
|-- vals: vector (nullable = true)
|-- hashValues: array (nullable = true)
| |-- element: vector (containsNull = true)
scala> df_transformed.show()
+------------+--------------------+--------------------+
| id| vals| hashValues|
+------------+--------------------+--------------------+
|401310732094|[-0.37154,-0.1159...|[[-949518.0], [47...|
|292125586474|[-0.30407,0.35437...|[[-764013.0], [31...|
|362051108485|[-0.36748,0.05738...|[[-688834.0], [18...|
|222480119030|[-0.2509,0.55574,...|[[-1167047.0], [2...|
|182270925238|[0.32288,-0.60789...|[[-836660.0], [97...|
+------------+--------------------+--------------------+
For example, I need to extract the value of the hashValues column into a DenseVectorfor id 401310732094.
This can be done with an UDF:
import spark.implicits._
val convertToVec = udf((array: Seq[Vector]) =>
Vectors.dense(array.flatMap(_.toArray).toArray)
)
val df = df_transformed.withColumn("hashValues", convertToVec($"hashValues"))
This will overwrite the hashValues column with a new one containing a DenseVector.
Tested with a dataframe with following schema:
root
|-- id: integer (nullable = false)
|-- hashValues: array (nullable = true)
| |-- element: vector (containsNull = true)
The result is:
root
|-- id: integer (nullable = false)
|-- hashValues: vector (nullable = true)
Related
I have a Spark DataFrame testDF with the following structure
scala> testDF.printSchema()
-------------------------------------------------
root
|-- id: long (nullable = true)
|-- array1: array (nullable = true)
| |-- element: integer (containsNull = true)
|-- array2: array (nullable = true)
| |-- element: integer (containsNull = true)
Both array1 and array2 are guaranteed to be of same length.
I want to perform bitwise xor between each element from array1 and array2.
Something like:
res = array1[0] ^ array2[0] + array1[1] ^ array2[1] + ...
I know this can be done using udf but wondering if there is a native spark way to do it.
Desired DataFrame Structure would look something like
root
|-- id: long (nullable = true)
|-- array1: array (nullable = true)
| |-- element: integer (containsNull = true)
|-- array2: array (nullable = true)
| |-- element: integer (containsNull = true)
|-- result: long (nullable = true)
The idea is to read a parquet file into dataFrame. Then, extract all column name's and type's from it's schema. If we have a nested columns, i would like to add a "prefix" before the column name.
Considering that we can have a nested column with sub column named properly, and we can have also a nested column with just an array of array without column name but "element".
val dfSource: DataFrame = spark.read.parquet("path.parquet")
val dfSourceSchema: StructType = dfSource.schema
Example of dfSourceSchema (Input):
|-- exCar: array (nullable = true)
| |-- element: array (containsNull = true)
| | |-- element: binary (nullable = true)
|-- exProduct: string (nullable = true)
|-- exName: array (nullable = true)
| |-- element: struct (containsNull = true)
| | |-- exNameOne: string (nullable = true)
| | |-- exNameTwo: string (nullable = true)
Desired output :
((exCar.prefix.prefix,binary)),(exProduct, String), (exName.prefix.exNameOne, String), (exName.prefix.exNameTwo, String) )
Spark 2 is converting scala array to WrappedArray automatically When i am passing array to function. However, In Spark 1.6 array is converted to string like '[a,b,c]' . Here is my code
val df_date_agg = df
.groupBy($"a",$"b",$"c")
.agg(sum($"d").alias("data1"),sum($"e").alias("data2"))
.groupBy($"a")
.agg(collect_list(array($"b",$"c",$"data1")).alias("final_data1"),
collect_list(array($"b",$"c",$"data2")).alias("final_data2"))
When I am running above code to spark 1.6. I am getting below schema
|-- final_data1: array (nullable = true)
| |-- element: string (containsNull = true)
|-- final_data2: array (nullable = true)
| |-- element: string (containsNull = true)
but in spark 2
|-- final_data1: array (nullable = true)
| |-- element: array (containsNull = true)
| | |-- element: string (containsNull = true)
|-- final_data1: array (nullable = true)
| |-- element: array (containsNull = true)
| | |-- element: string (containsNull = true)
How can I change datatype of spark 2 as per spark 1?
Since you want a string representation of an array, how about casting the array into a string like this?
val df_date_agg = df
.groupBy($"a",$"b",$"c")
.agg(sum($"d").alias("data1"),sum($"e").alias("data2"))
.groupBy($"a")
.agg(collect_list(array($"b",$"c",$"data1") cast "string").alias("final_data1"),
collect_list(array($"b",$"c",$"data2") cast "string").alias("final_data2"))
It might simply be what your old version of spark was doing. I was not able to verify.
my DataFrame like this :
+------------------------+----------------------------------------+
|ID |probability |
+------------------------+----------------------------------------+
|583190715ccb64f503a|[0.49128147201958017,0.5087185279804199]|
|58326da75fc764ad200|[0.42143416087939345,0.5785658391206066]|
|583270ff17c76455610|[0.3949217100212508,0.6050782899787492] |
|583287c97ec7641b2d4|[0.4965059792664432,0.5034940207335569] |
|5832d7e279c764f52e4|[0.49128147201958017,0.5087185279804199]|
|5832e5023ec76406760|[0.4775830044196701,0.52241699558033] |
|5832f88859cb64960ea|[0.4360509428173421,0.563949057182658] |
|58332e6238c7643e6a7|[0.48730029128352853,0.5126997087164714]|
and I get the column of probability using
val proVal = Data.select("probability").rdd.map(r => r(0)).collect()
proVal.foreach(println)
the result is :
[0.49128147201958017,0.5087185279804199]
[0.42143416087939345,0.5785658391206066]
[0.3949217100212508,0.6050782899787492]
[0.4965059792664432,0.5034940207335569]
[0.49128147201958017,0.5087185279804199]
[0.4775830044196701,0.52241699558033]
[0.4360509428173421,0.563949057182658]
[0.48730029128352853,0.5126997087164714]
but I want to get the first column of data for each row, like this:
0.49128147201958017
0.42143416087939345
0.3949217100212508
0.4965059792664432
0.49128147201958017
0.4775830044196701
0.4360509428173421
0.48730029128352853
how can this be done?
The input is standard random forest input, above the input is val Data = predictions.select("docID", "probability")
predictions.printSchema()
root
|-- docID: string (nullable = true)
|-- label: double (nullable = false)
|-- features: vector (nullable = true)
|-- indexedLabel: double (nullable = true)
|-- rawPrediction: vector (nullable = true)
|-- probability: vector (nullable = true)
|-- prediction: double (nullable = true)
|-- predictedLabel: string (nullable = true)
and I want to get the first value of the "probability" column
You can use the Column.apply method to get the n-th item on an array column - in this case the first column (using index 0):
import sqlContext.implicits._
val proVal = Data.select($"probability"(0)).rdd.map(r => r(0)).collect()
BTW, if you're using Spark 1.6 or higher, you can also use the Dataset API for a cleaner way to convert the dataframe into Doubles:
val proVal = Data.select($"probability"(0)).as[Double].collect()
I have two dataframes df1, df2 whose schema is as follows:
DF1 is of the form:
slotSize2: struct (nullable = true)
| |-- 120x600: struct (nullable = true)
| | |-- pView: string (nullable = true)
| |-- 160x600: struct (nullable = true)
| | |-- level: string (nullable = true)
| | |-- pATF: string (nullable = true)
| | |-- pView: string (nullable = true)
| | |-- pViewV1: string (nullable = true)
| | |-- sPos: string (nullable = true)
| |-- 250x250: struct (nullable = true)
| | |-- pView: string (nullable = true)
| |-- 300x250: struct (nullable = true)
| | |-- level: string (nullable = true)
| | |-- pATF: string (nullable = true)
| | |-- pView: string (nullable = true)
| | |-- pViewV1: string (nullable = true)
| | |-- sPos: string (nullable = true)
Dataframe df2 has schema :
root
|-- bidId: array (nullable = true)
| |-- element: string (containsNull = true)
|-- slotSize1: array (nullable = true)
| |-- element: string (containsNull = true)
In dataframe df2 we have slotSize (named as slotSize1) in form of string and in dataframe df1 we have nested form of slotsize i.e. for every slotsize we have corresponding map.
I want to join two dataframes df1, df2 to form a new dataframe df3 which has schema (bidId, slotSize , viewMap) where bidId is present in df1 , slotSIze is of form 120x600 and is present in both schemas, and viewMap corresponds to the nested map corresponding to every slotSize in df1.