[ Skip to the content ]

Institute of Formal and Applied Linguistics Wiki

[ Back to the navigation ]

You are here: start » spark » using-python

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
Next revision 		Previous revision
Last revision Both sides next revision
spark:using-python [2014/11/10 15:36]
straka 		spark:using-python [2022/12/14 13:25]
straka [Usage Examples]
Line 8: Line 8:
 <file>pyspark</file> <file>pyspark</file>
  
-Better interactive shell with code completion using ''ipython'' (installed everywhere on cluster; ask our IT if you want to have it installed on your workstations too) can be started using: +Better interactive shell with code completion using ''ipython3'' (do `pip3 install --user ipython` if you do not have it) can be started using: 
-<file>IPYTHON=1 pyspark</file>+<file>PYSPARK_DRIVER_PYTHON=ipython3 pyspark</file>
  
 As described in [[running-spark-on-single-machine-or-on-cluster|Running Spark on Single Machine or on Cluster]], environmental variable ''MASTER'' specifies which Spark master to use (or whether to start a local one). As described in [[running-spark-on-single-machine-or-on-cluster|Running Spark on Single Machine or on Cluster]], environmental variable ''MASTER'' specifies which Spark master to use (or whether to start a local one).
Line 16: Line 16:
 Consider the following simple script computing 10 most frequent words of Czech Wikipedia: Consider the following simple script computing 10 most frequent words of Czech Wikipedia:
 <file python> <file python>
-(sc.textFile("/net/projects/spark-example-data/wiki-cs")+(sc.textFile("/net/projects/spark-example-data/wiki-cs", 3*sc.defaultParallelism)
    .flatMap(lambda line: line.split())    .flatMap(lambda line: line.split())
    .map(lambda word: (word, 1))    .map(lambda word: (word, 1))
-   .reduceByKey(lambda c1,c2: c1+c2) +   .reduceByKey(lambda c1, c2: c1+c2) 
-   .sortBy(lambda (word,count): count, ascending=False)+   .sortBy(lambda word_count: word_count[1], ascending=False)
    .take(10))    .take(10))
 </file> </file>
  
-  * run interactive shell inside ''spark-qrsh'', or start local Spark cluster using as many threads as there are cores: +  * run interactive shell using existing Spark cluster (i.e., inside ''spark-srun''), or start local Spark cluster using as many threads as there are cores if there is none: 
-  <file>IPYTHON=1 pyspark</file>+  <file>PYSPARK_DRIVER_PYTHON=ipython3 pyspark</file>
   * run interactive shell with local Spark cluster using one thread:   * run interactive shell with local Spark cluster using one thread:
-  <file>MASTER=local IPYTHON=1 pyspark</file> +  <file>MASTER=local PYSPARK_DRIVER_PYTHON=ipython3 pyspark</file> 
-  * start Spark cluster (10 machines, 1GB RAM each) on SGE and run interactive shell: +  * start Spark cluster (10 machines, 2GB RAM each) on Slurm and run interactive shell: 
-  <file>IPYTHON=1 spark-qrsh 10 1G pyspark</file>+  <file>PYSPARK_DRIVER_PYTHON=ipython3 spark-srun 10 2G pyspark</file>
  
-Note that ''IPYTHON'' variable can be left out or specified in ''.bashrc'' (or similar).+Note that ''PYSPARK_DRIVER_PYTHON'' variable can be left out or specified in ''.bashrc'' (or other configuration files).
  
  
Line 56: Line 56:
        
 sc = SparkContext() sc = SparkContext()
-(sc.textFile(input)+(sc.textFile(input, 3*sc.defaultParallelism)
    .flatMap(lambda line: line.split())    .flatMap(lambda line: line.split())
    .map(lambda token: (token, 1))    .map(lambda token: (token, 1))
-   .reduceByKey(lambda x,y: x + y) +   .reduceByKey(lambda x, y: x + y) 
-   .sortBy(lambda (word,count): count, ascending=False)+   .sortBy(lambda word_count: word_count[1], ascending=False)
    .saveAsTextFile(output))    .saveAsTextFile(output))
 sc.stop() sc.stop()
 </file> </file>
  
-  * run ''word_count.py'' script inside ''spark-qsub'', ''spark-qrsh'', or start local Spark cluster using as many threads as there are cores: +  * run ''word_count.py'' script inside existing Spark cluster (i.e., inside ''spark-sbatch'' or ''spark-srun''), or start local Spark cluster using as many threads as there are cores if there is none: 
-  <file>spark-submit word_count.py input output</file>+  <file>spark-submit word_count.py /net/projects/spark-example-data/wiki-cs outdir</file>
   * run ''word_count.py'' script with local Spark cluster using one thread:   * run ''word_count.py'' script with local Spark cluster using one thread:
-  <file>MASTER=local spark-submit word_count.py input output</file> +  <file>MASTER=local spark-submit word_count.py /net/projects/spark-example-data/wiki-cs outdir</file> 
-  * start Spark cluster (10 machines, 1GB RAM each) on SGE and run ''word_count.py'' script: +  * start Spark cluster (10 machines, @GB RAM each) using Slurm SGE and run ''word_count.py'' script: 
-  <file>spark-qsub 10 1G spark-submit word_count.py input output</file>+  <file>spark-sbatch 10 2G spark-submit word_count.py /net/projects/spark-example-data/wiki-cs outdir</file> 
 + 
 +===== Using Virtual Environments ===== 
 + 
 +If you want to use specific virtual environment in your Spark job, use 
 +<file>PYSPARK_PYTHON=path_to_python_in_venv [pyspark|spark-submit]</file> 

[ Back to the navigation ] [ Back to the content ]