Editing CommunityData:Hyak Spark

If you are using the [[CommunityData:Hyak-Mox | cdsc mox setup]] then you should have a working spark configuration in your environment already.  Otherwise, you'll need to have the following in your .bashrc (remember to source .bashrc or re-login to load up these changes):

     export JAVA_HOME='/com/local/java/'

     export PATH="$JAVA_HOME/bin:$PATH"

     export TMPDIR="/com/users/[YOU]/tmpdir"

You can also run spark programs on many nodes, but this requires additional steps. These are described below.

The script is on ikt here: /com/users/nathante/mediawiki_dump_tools/wikiq_users/wikiq_users_spark.py

     #!/usr/bin/env python3                                                                                                                                                                                             

     import sys

     from os import mkdir

     from os import path

This part imports some python utilities that we will use.  You can pretty safely treat the <code>SparkConf</code>, the <code>SparkSession</code> and the <code>SQLContext</code> imports as magic that creates a spark environment that supports working with Spark's SQL features.  

<code>Window</code> is used to create Window functions. We will use a window function to count the nth edit made by each editor. <code>import pyspark.sql.functions as f</code> provides built in functions that can be applied to data in spark data frames. <code>types</code> are data types that we will use to specify the scheme for reading wikiq files.  

     def parse_args():

         parser = argparse.ArgumentParser(description='Create a dataset of edits by user.')

         args = parser.parse_args()

         return(args)

Above is just a function to build a command line interface.  

     if __name__ == "__main__":

         conf = SparkConf().setAppName("Wiki Users Spark")

         spark = SparkSession.builder.getOrCreate()

Now we are in the main function of the script. The above two lines complete setting up spark. If you are going to run this program on a multi-node cluster, then it would be nice to set the AppName to something friendly. This will be used by the job monitoring tools.   

         args = parse_args()

         files = glob.glob(args.input_file)

         files = [path.abspath(p) for p in files]

   

Spark is designed to read and write lists of files. The <code>args.input_file</code> uses <code>glob</code> to accept wildcards. The above lines build a list of files from the argument.  

     reader = spark.read

This creates a reader object that can read files. We are starting to get down to business. Next we will specify the schema for the files that we will read in. This is important so that spark can run efficiently and operate on the correct data types.  

     # build a schema                                                                                                                                                                                               

     struct = types.StructType().add("anon",types.StringType(),True)

     struct = struct.add("text_chars", types.LongType(), True)

     struct = struct.add("title",types.StringType(), True)

This is a little bit tedious, but it is necessary for Spark to work effectively on tsv data. If you are reading binary format such as Parquet (which is recommended, and easy to create using Spark) then you can skip this.  

     df = reader.csv(files,

                     sep='\t',

                     mode="PERMISSIVE",

                     schema = struct)

This reads the data into a Spark Dataframe. Spark Dataframes are more like SQL tables than they are like pandas DataFrames.  Spark Dataframes are pretty abstract and can live on memory or on disk. Operations on Spark Dataframes are lazily evaluated, Spark will not actually run computations on your data until it has to. Calling <code>df.show()</code> will print the dataframe and trigger execution. <code>mode="PERMISSIVE"</code> stops Spark from giving up if it hits malformed rows.  

     df = df.repartition(args.num_partitions)

The first thing to do after reading the data is to <code>repartition</code> the data. This determines the number of files that spark will output. Choosing the right number of partitions isn't really an exact science. Having more partitions makes some operations more efficient and can make other operations slower. The rule of thumb is that the number of partitions increases linearly with the amount of data. 500 partitions seems pretty good for English Wikipedia. If you are interested [https://jaceklaskowski.gitbooks.io/mastering-apache-spark/content/spark-rdd-partitions.html this page] is good. Now we are ready to build some variables. The first thing we are going to do is to create a new column <code>editor_id_or_ip</code>.

     # replace na editor ids                                                                                                                                                                                         

     df = df.select('*',f.coalesce(df['editor_id'],df['editor']).alias('editor_id_or_ip'))

The first argument to select <code>'*'</code> causes select to return all the columns.  Next we call the <code>coalesce</code> function which creates a new column with the value of <code>editor_id</code> if <code>editor_id</code> is not null and the value of <code>editor</code> if <code>editor_id</code> is null. The call to <code>alias</code> gives the new column a name.  If you are familiar with SQL programming, this might seem familiar. You could write it as <code>SELECT *, COALESCE(editor_id, editor) AS editor_id_or_ip</code>.  

Next we are going to identify the edits that revert each edit.  <code>reverteds</code> lists the edits that the edit has reverted.

     # assign which edit reverted what edit                                                                                                                                                                         

     reverteds_df = df.filter(~ df.reverteds.isNull()).select(['revid','reverteds'])

This line creates a new Spark Dataframe out of the rows of the first dataframe that have a value for <code>reverteds</code> with the columns <code>revid</code> and <code>reverteds</code>.  

     reverteds_df = reverteds_df.select("*", f.split(reverteds_df.reverteds,',').alias("reverteds_new"))

The above line converts <code>reverteds</code> from a string to an array.  

     reverteds_df = reverteds_df.drop("reverteds")

     reverteds_df = reverteds_df.withColumnRenamed("reverteds_new", "reverteds")

The above two lines remove the old "reverteds" column, which was a string, and replaces it with the array column. This is required because unlike pandas, Spark dataframes do not have a column assignment syntax.   

     reverteds_df = reverteds_df.select(reverteds_df.revid.alias('reverted_by'),

                                       f.explode(reverteds_df.reverteds).alias('reverted_id'))

The most important part of the above is the function call to <code>explode</code>. Explode💥 unfolds the array so that we get one line for each element of the array. Now we can join <code>reverteds_df</code> with <code>df</code> to put the <code>reverted_by</code> column in <code>df</code>.

     df = df.join(reverteds_df, df.revid == reverteds_df.reverted_id, how='left_outer')

     df.drop("reverted_id")

     del(reverteds_df)

Join the two tables so that each revision that was reverted gets a value for <code>reverted_by</code>. There are many kinds of joins and there is some detail on this in the [[#Join help|Join help section of this page]]. The join is a <code>left_outer</code> join so we keep all the rows of <code>df</code> even the rows that don't have a value for <code>reverted_id</code> in <code>reverteds_df</code>. We remove the redundent <code>reverted_id</code> column and are don with building <code>reverted_by</code>. Next we add a column that counts the number of times a given editor has made a revert (this is called a ''cumulative count''). Since we aren't going to use <code>reverteds_df</code> again we can call <code>del(reveteds_df)</code>. This tells spark it is free to remove the object from storage and can improve performance.  

     # sort by datetime                                                                                                                                                                                             

     df = df.orderBy(df.date_time.asc())

<code>orderBy</code> sorts the dataframe by date.   

     win = Window.orderBy('date_time').partitionBy('editor_id_or_ip')

The above defines a <code>WindowSpec</code>, which is a kind of object that can be used to define rolling aggregations. We are going to use the <code>rank</code> function to perform the cumulative count, and <code>rank</code> requires a <code>WindowSpec</code>. The WindowSpec that we made says that we are grouping at the level of <code>editor_id_or_ip</code> and that we want to operate on each row of each group in chronological order.  

     # count reverts                                                                                                                                                                                                 

     reverts_df = df.filter(df.revert==True).select(['revid','editor_id_or_ip','date_time','revert'])

The above creates a new table that only has reverts.   

     reverts_df = reverts_df.withColumn('editor_nth_revert',f.rank().over(win))

This applies the <code>rank</code> function over the window to perform the cumulative count of the reverts. The <code>withColumn</code> function adds a new column to the dataframe called <code>editor_nth_revert</code>.

     df = df.join(reverts_df, ["revid",'editor_id_or_ip','date_time','revert'], how='left_outer')

     del(reverts_df)

Above we perform the join to add the new column to <code>df</code>. We join on all of the columns <code> ["revid",'editor_id_or_ip','date_time','revert']</code> so that duplicate columns are not created in the <code>df</code>

     # count edits                                                                                                                                                                                                   

     df = df.withColumn('year', f.year(df.date_time))

     df = df.withColumn('month',f.month(df.date_time))

Using <code>withColumn</code> again to illustrate creating some calendar variables from the <code>date_time</code>.  

     df = df.withColumn('editor_nth_edit',f.rank().over(win))

We can reuse the <code>WindowSpec</code> to get the cumulative count for all edits as opposed to all reverts.  

     # output                                                                                                                                                                                                       

     if not path.exists(args.output_dir):

     if args.output_format == "csv" or args.output_format == "tsv":

         df.write.csv(args.output_dir, sep='\t', mode='overwrite',header=True,timestampFormat="yyyy-MM-dd HH:mm:ss")

Instead of writing our output to a single file, we output to a directory. Spark will write 1 file for each partition to the directory.  

             # format == "parquet"                                                                                                                                                                                   

     else:

         df.write.parquet(args.output_dir, mode='overwrite')

It is also easy to write to parquet.