In a Spark (2.3.0) project using Scala, I would like to drop multiple columns using a regex. I tried using colRegex, but without success:
val df = Seq(("id","a_in","a_out","b_in","b_out"))
.toDF("id","a_in","a_out","b_in","b_out")
val df_in = df
.withColumnRenamed("a_in","a")
.withColumnRenamed("b_in","b")
.drop(df.colRegex("`.*_(in|out)`"))
// Hoping to get columns Array(id, a, b)
df_in.columns
// Getting Array(id, a, a_out, b, b_out)
On the other hand, the mechanism seems to work with select:
df.select(df.colRegex("`.*_(in|out)`")).columns
// Getting Array(a_in, a_out, b_in, b_out)
Several things are not clear to me:
what is this backquote syntax in the regex?
colRegex returns a Column: how can it actually represent several columns in the 2nd example?
can I combine drop and colRegex or do I need some workaround?
If you check spark code of colRefex method ... it expects regexs to be passed in the below format
/** the column name pattern in quoted regex without qualifier */
val escapedIdentifier = "`(.+)`".r
/** the column name pattern in quoted regex with qualifier */
val qualifiedEscapedIdentifier = ("(.+)" + """.""" + "`(.+)`").r
backticks(`) are necessary to enclose your regex, otherwise the above patterns will not identify your input pattern.
you can try selecting specific colums which are valid as mentioned below
val df = Seq(("id","a_in","a_out","b_in","b_out"))
.toDF("id","a_in","a_out","b_in","b_out")
val df_in = df
.withColumnRenamed("a_in","a")
.withColumnRenamed("b_in","b")
.drop(df.colRegex("`.*_(in|out)`"))
val validColumns = df_in.columns.filter(p => p.matches(".*_(in|out)$")).toSeq //select all junk columns
val final_df_in = df_in.drop(validColumns:_*) // this will drop all columns which are not valid as per your criteria.
In addition to the workaround proposed by Waqar Ahmed and kavetiraviteja (accepted answer), here is another possibility based on select with some negative regex magic. More concise, but harder to read for non-regex-gurus...
val df_in = df
.withColumnRenamed("a_in","a")
.withColumnRenamed("b_in","b")
.select(df.colRegex("`^(?!.*_(in|out)_).*$`")) // regex with negative lookahead
Related
I have a scenario where I need to form a dataframe name from two string variable. Which is pretty easy and can be done by concatenating.
Example: "df_" + "part1324"
The above code will return a String variable. I want this to be a Dataframe variable through which I can perform further operation on the data frame.
Map can be used for assign names to DataFrames:
val df = List(("df_value")).toDF()
val stringVariable = "part1324"
// assign name to dataframe
val namedDataFrames = Map("df_" + stringVariable -> df)
// get dataframe by name
namedDataFrames("df_part1324").show(false)
Your question is confusing. What do you mean by dataframe variable? Concatenating two strings will always return String. In order to create a dataframe, you need to apply the different methods available to create a dataframe.
val df:Dataframe cannot be equal to df_part1234 (String)as per your example but to use it as dataframe, you need to do something like below
val df_part1234 = sc.range(1000).toDF("number") where sc is your Sparksession variable.
In case you need to generate this variable dynamically, place it under the logic of variable generation like Loop and add the statement to create the dataframe.
Please rewrite your question if you are trying to achieve something else (along with code snippet to reproduce the issue) or accept the answer if you are clear on the issue
I see a code from book "Spark The Definitive Guide",it invoke a drop on a dataframe with no parameter,when I use show(),I found nothing changed,but what is the meaning of it?
I execute it,nothing changed,dfNoNull.show() is the same as dfWithDate.show()
dfWithDate.createOrReplaceTempView("dfWithDate")
// in Scala
val dfNoNull = dfWithDate.drop()
dfNoNull.createOrReplaceTempView("dfNoNull")
does it mean, it create a new datarframe?
I know when a dataframe join itself when I using Hive sql,if I just
val df1=spark.sql("select id,date from date")
val df2=spark.sql("select id,date from date")
val joinedDf = spark.sql("select dateid1,dateid2 from sales")
.join(df1,df1["id"]===dateid1).join(df2,df2["id"]===dateid2)
Then an error occur:Cartesian join!
because the lazy evalution will consider df1 and df1 as the same one
so here,if I
val df2=df1.drop()
will I prevent that error?
If not,what does the drop method with no parameter mean?
Or it just mean remove the temp view name and create a new one?
but I try the code below,no exception throwed:
val df= Seq((1,"a")).toDF("id","name")
df.createOrReplaceTempView("df1")
val df2=df.drop()
df2.createOrReplaceTempView("df2")
spark.sql("select * from df1").show()
Or does the book mean below?
val dfNoNull = dfWithDate.na.drop()
because it wrote somewhere below the code:
Grouping sets depend on null values for aggregation levels. If you do
not filter-out null values, you will get incorrect results.This
applies to cubes, rollups, and grouping sets.
drop function with no parameter behave the same as drop with column name that doesn't exist in the Dataframe.
You can follow the code in the source of spark.
Even in the function documentation you can see a hint to this behavior.
/**
* Returns a new Dataset with a column dropped. This is a no-op if schema doesn't contain
* column name.
*
* This method can only be used to drop top level columns. the colName string is treated
* literally without further interpretation.
*
* #group untypedrel
* #since 2.0.0
*/
So when calling the function with no parameter no-op occur and nothing changes in the returning DataFrame.
I have a data frame with n number of columns and I want to replace empty strings in all these columns with nulls.
I tried using
val ReadDf = rawDF.na.replace("columnA", Map( "" -> null));
and
val ReadDf = rawDF.withColumn("columnA", if($"columnA"=="") lit(null) else $"columnA" );
Both of them didn't work.
Any leads would be highly appreciated. Thanks.
Your first approach seams to fail due to a bug that prevents replace from being able to replace values with nulls, see here.
Your second approach fails because you're confusing driver-side Scala code for executor-side Dataframe instructions: your if-else expression would be evaluated once on the driver (and not per record); You'd want to replace it with a call to when function; Moreover, to compare a column's value you need to use the === operator, and not Scala's == which just compares the driver-side Column object:
import org.apache.spark.sql.functions._
rawDF.withColumn("columnA", when($"columnA" === "", lit(null)).otherwise($"columnA"))
I would like to aggregate a Spark data frame using an array of column names as input, and at the same time retain the original names of the columns.
df.groupBy($"id").sum(colNames:_*)
This works but fails to preserve the names. Inspired by the answer found here I unsucessfully tried this:
df.groupBy($"id").agg(sum(colNames:_*).alias(colNames:_*))
error: no `: _*' annotation allowed here
It works to take a single element like
df.groupBy($"id").agg(sum(colNames(2)).alias(colNames(2)))
How can make this happen for the entire array?
Just provide an sequence of columns with aliases:
val colNames: Seq[String] = ???
val exprs = colNames.map(c => sum(c).alias(c))
df.groupBy($"id").agg(exprs.head, exprs.tail: _*)
I am new to Spark. I have two tables in HDFS. One table(table 1) is a tag table,composed of some text, which could be some words or a sentence. Another table(table 2) has a text column. Every row could have more than one keyword in the table 1. my task is find out all the matched keywords in table 1 for the text column in table 2, and output the keyword list for every row in table 2.
The problem is I have to iterate every row in table 2 and table 1. If I produce a big list for table 1, and use a map function for table 2. I will still have to use a loop to iterate the list in the map function. And the driver shows the JVM memory limit error,even if the loop is not large(10 thousands time).
myTag is the tag list of table 1.
def ourMap(line: String, myTag: List[String]): String = {
var ret = line
val length = myTag.length
for (i <- 0 to length - 1) {
if (line.contains(myTag(i)))
ret = ret.replaceAll(myTag(i), "_")
}
ret
}
val matched = result.map(b => ourMap(b, tagList))
Any suggestion to finish this task? With or without Spark
Many thanks!
An example is as follows:
table1
row1|Spark
row2|RDD
table2
row1| Spark is a fast and general engine. RDD supports two types of operations.
row2| All transformations in Spark are lazy.
row3| It is for test. I am a sentence.
Expected result :
row1| Spark,RDD
row2| Spark
MAJOR EDIT:
The first table actually may contain sentences and not just simple keywords :
row1| Spark
row2| RDD
row3| two words
row4| I am a sentence
Here you go, considering the data sample that you have provided :
val table1: Seq[(String, String)] = Seq(("row1", "Spark"), ("row2", "RDD"), ("row3", "Hashmap"))
val table2: Seq[String] = Seq("row1##Spark is a fast and general engine. RDD supports two types of operations.", "row2##All transformations in Spark are lazy.")
val rdd1: RDD[(String, String)] = sc.parallelize(table1)
val rdd2: RDD[(String, String)] = sc.parallelize(table2).map(_.split("##").toList).map(l => (l.head, l.tail(0))).cache
We'll build an inverted index of the second data table which we will join to the first table :
val df1: DataFrame = rdd1.toDF("key", "value")
val df2: DataFrame = rdd2.toDF("key", "text")
val df3: DataFrame = rdd2.flatMap { case (row, text) => text.trim.split( """[^\p{IsAlphabetic}]+""")
.map(word => (word, row))
}.groupByKey.mapValues(_.toSet.toSeq).toDF("word", "index")
import org.apache.spark.sql.functions.explode
val results: RDD[(String, String)] = df3.join(df1, df1("value") === df3("word")).drop("key").drop("value").withColumn("index", explode($"index")).rdd.map {
case r: Row => (r.getAs[String]("index"), r.getAs[String]("word"))
}.groupByKey.mapValues(i => i.toList.mkString(","))
results.take(2).foreach(println)
// (row1,Spark,RDD)
// (row2,Spark)
MAJOR EDIT:
As mentioned in the comment : The specifications of the issue changed. Keywords are no longer simple keywords, they might be sentences. In that case, this approach wouldn't work, it's a different kind of problem. One way to do it is using Locality-sensitive hashing (LSH) algorithm for nearest neighbor search.
An implementation of this algorithm is available here.
The algorithm and its implementation are unfortunately too long to discuss on SO.
From what I could gather from your problem statement is that you are kind of trying to tag the data in Table 2 with the keywords which are present in Table 1. For this, instead of loading the Table1 as a list and then doing each keyword pattern matching for each row in Table2, do this :
Load Table1 as a hashSet.
Traverse the Table2 and for each word in that phrase, do a search in the above hashset. I assume the words that you shall have to search from here are less as compared to pattern matching for each keyword. Remember, search now is O(1) operation whereas pattern matching is not.
Also, in this process, you can also filter words like " is, are, when, if " etc as they shall never be used for tagging. So that reduces words you need to find in hashSet.
The hashSet can be loaded into memory(I think 10K keywords should not take more than few MBs). This variable can be shared across executors through broadcast variables.