How to retrieve nothing out of a spark dataframe.
I need something like this,
df.where("1" === "2")
I needed this so that I can do a left join with another dataframe.
Basically I am trying to avoid the data skewing while joining two dataframes by splitting the null and not null key columns and joining them separately and then do a union them.
df1 has 300M records out of which 200M records has Null keys.
df2 has another 300M records.
So to join them, I am splitting the df1 containing null and not null keys separately and then join them with df2. so to join the null key dataframe with df2, I don't need any records from df2.
I can just add the columns from df2 to null key df1,
but curious to see if we have something like this in spark
df.where("1" === "2")
As we do in RDBMS SQLs.
There many different ways, like limit:
df.limit(0)
where with Column:
import org.apache.spark.sql.functions._
df.where(lit(false))
where with String expression:
df.where("false")
1 = 2 expressed as
df.where("1 = 2")
or
df.where(lit(1) === lit(2))
would work as well, but are more verbose than required.
where function calls filter function at the internal level so you can use filter as
import org.apache.spark.sql.functions._
df.filter(lit(1) === lit(2))
or
import org.apache.spark.sql.functions._
df.filter(expr("1 = 2"))
or
df.filter("1 = 2")
or
df.filter("false")
or
import org.apache.spark.sql.functions._
df.filter(lit(false))
Any expression that would return false in the filter function would work.
Related
I have a scenario here - Have 30 columns in one dataframe, need to drop specific column and select remaining columns and put it to another dataframe. How can I achieve this? I tried below.
val df1: DataFrame = df2.as(a).join( df3.as(b),col(a.key) === col(b.key), inner).drop(a.col1)
.select(" a.star ")
when I do show of df1, its still show col1. Any advise on resolving this.
drop requires a string without table alias, so you can try:
val df1 = df2.as("a")
.join(df3.as("b"), col("a.key") === col("b.key"), "inner")
.drop("col1")
.select("a.*")
Or instead of dropping the column, you can filter the columns to be selected:
val df1 = df2.as("a")
.join(df3.as("b"), col("a.key") === col("b.key"), "inner")
.select(df2.columns.filterNot(_ == "col1").map("a." + _): _*)
This really just seems like you need to use a "left_semi" join.
val df1 = df2.drop('col1).join(df3, df2("key") === df3("key"), "left_semi")
If key is an actual column you can simplify the syntax even further
val df1 = df2.drop('col1).join(df3, Seq("key"), "left_semi")
The best syntax depends on the details of what your real data looks like. If you need to refer to col1 in df2 specifically because there's ambiguity, then use df2("col1")
left_semi joins take all the columns from the left table for rows finding a match in the right table.
I have a Dataframe resulting from a join of two Dataframes: df1 and df2 into df3. All the columns found in df2 are also in df1, but their contents differ. I'd like to remove all the df1 columns which names are in df2.columns from the join. Would there be a way to do this without using a var?
Currently I've done this
var ret = df3
df2.columns.foreach(coln => ret = ret.drop(df2(coln)))
but what I really want is just a shortcut for
df3.drop(df1(df2.columns(1))).drop(df1(df2.columns(2)))....
without using a var.
Passing a list of columns is not an option, don't know if it's because I'm using spark 2.2
EDIT:
Important note: I don't know in advance the columns of df1 and df2
This is possible to achieve while you are performing the join itself. Please try the below code
val resultDf=df1.alias("frstdf").join(broadcast(df2).alias("scndf"), $"frstdf.col1" === $"scndf.col1", "left_outer").selectExpr("scndf.col1","scndf.col2"...)//.selectExpr("scndf.*")
This would only contain the columns from the second data frame. Hope this helps
A shortcut would be:
val ret = df2.columns.foldLeft(df3)((acc,coln) => acc.drop(df2(coln)))
I would suggest to remove the columns before the join. Alternatively, select only the columns from df3 which come from df2:
val ret = df3.select(df2.columns.map(col):_*)
I have 2 dataframes : df1 and df2 and I am left joining both of them on id column and saving it to another dataframe named df3. Below is the code that I am using, which works fine as expected.
val df3 = df1.alias("tab1").join(df2.alias("tab2"),Seq("id"),"left_outer").select("tab1.*","tab2.name","tab2.dept","tab2.descr");
I would like to rename the tab2.descr column to dept_full_description within the above statement.
I am aware that I could create a seq val like below and use toDF method
val columnsRenamed = Seq("id", "empl_name", "name","dept","dept_full_description") ;
df4 = df3.toDF(columnsRenamed: _*);
Is there any other way to to aliasing in the first statement itself. My end goal is not to list about 30-40 columns explicitly .
I'd rename before join:
df1.alias("tab1").join(
df2.withColumnRenamed("descr", "dept_full_description").alias("tab2"),
Seq("id"), "left_outer")
I have to join two dataframes, which is very similar to the task given here Joining two DataFrames in Spark SQL and selecting columns of only one
However, I want to select only the second column from df2. In my task, I am going to use the join function for two dataframes within a reduce function for a list of dataframes. In this list of dataframes, the column names will be different. However, in each case I would want to keep the second column of df2.
I did not find anywhere how to select a dataframe's column by their numbered index. Any help is appreciated!
EDIT:
ANSWER
I figured out the solution. Here is one way to do this:
def joinDFs(df1: DataFrame, df2: DataFrame): DataFrame = {
val df2cols = df2.columns
val desiredDf2Col = df2cols(1) // the second column
val df3 = df1.as("df1").join(df2.as("df2"), $"df1.time" === $"df2.time")
.select($"df1.*",$"df2.$desiredDf2Col")
df3
}
And then I can apply this function in a reduce operation on a list of dataframes.
var listOfDFs: List[DataFrame] = List()
// Populate listOfDFs as you want here
val joinedDF = listOfDFs.reduceLeft((x, y) => {joinDFs(x, y)})
To select the second column in your dataframe you can simply do:
val df3 = df2.select(df2.columns(1))
This will first find the second column name and then select it.
If the join and select methods that you want to define in reduce function is similar to Joining two DataFrames in Spark SQL and selecting columns of only one Then you should do the following :
import org.apache.spark.sql.functions._
d1.as("d1").join(d2.as("d2"), $"d1.id" === $"d2.id").select(Seq(1) map d2.columns map col: _*)
You will have to remember that the name of the second column i.e. Seq(1) should not be same as any of the dataframes column names.
You can select multiple columns as well but remember the bold note above
import org.apache.spark.sql.functions._
d1.as("d1").join(d2.as("d2"), $"d1.id" === $"d2.id").select(Seq(1, 2) map d2.columns map col: _*)
Suppose there is one column in dataframe and there is similar schema column in another dataframe. how to check check the values consisting in the columns are same or not without joining them as there is not common attribute.
DF1
serial_nm
abc
mnc
pqr
DF2
ser_nm
hgf
mnc
uio
pqr
lok
And i want third DF3 as output
DF3
mnc
pqr
I tried this
val DF3 = DF1.filter(DF1("serial_nm") === DF2("ser_nm"))
But its not working
Please Help
Thanks..!!
I believe you can use a join. Consider using it like this:
val DF3 = DF1.join(DF2, DF1("serial_nm") === DF2("ser_nm"))
or
val DF3 = DF1.join(DF2).where(DF1("serial_nm") === DF2("ser_nm"))
Both approaches are quivalent.
Note: To avoid problems with ambiguous columns, one option is to rename them before the join:
val df2_renamed = DF2
.withColumnRenamed("mnc", "df2_mnc")
.withColumnRenamed("pqr", "df2_pqr")