This repository serves as a comprehensive guide to PySpark, featuring theory and exercises sourced from DataCamp. It is designed for beginners looking to understand the fundamentals of PySpark and its applications in big data processing. Here, you'll find detailed explanations, practical examples, and hands-on exercises to reinforce your learning. Whether you're just starting out or looking to enhance your skills, this repo provides valuable resources to help you master PySpark.
âś… Getting started with machine learning pipelines
âś… Model tuning and selection
Spark is a platform for cluster computing. Spark lets you spread data and computations over clusters with multiple nodes (think of each node as a separate computer). Splitting up your data makes it easier to work with very large datasets because each node only works with a small amount of data. As each node works on its own subset of the total data, it also carries out a part of the total calculations required, so that both data processing and computation are performed in parallel over the nodes in the cluster. It is a fact that parallel computation can make certain types of programming tasks much faster. However, with greater computing power comes greater complexity. Deciding whether or not Spark is the best solution for your problem takes some experience, but you can consider questions like:
- Is my data too big to work with on a single machine?
- Can my calculations be easily parallelized?
The first step in using Spark is connecting to a cluster.
In practice, the cluster will be hosted on a remote machine that's connected to all other nodes. There will be one computer, called the master that manages splitting up the data and the computations. The master is connected to the rest of the computers in the cluster, which are called worker. The master sends the workers data and calculations to run, and they send their results back to the master.
When you're just getting started with Spark it's simpler to just run a cluster locally. Thus, for this course, instead of connecting to another computer, all computations will be run on DataCamp's servers in a simulated cluster.
Creating the connection is as simple as creating an instance of the SparkContext class. The class constructor takes a few optional arguments that allow you to specify the attributes of the cluster you're connecting to.
An object holding all these attributes can be created with the SparkConf() constructor. Take a look at the documentation for all the details!
For the rest of this course you'll have a SparkContext called sc already available in your workspace.
In this exercise you'll get familiar with the SparkContext.
You'll probably notice that code takes longer to run than you might expect. This is because Spark is some serious software. It takes more time to start up than you might be used to. You may also find that running simpler computations might take longer than expected. That's because all the optimizations that Spark has under its hood are designed for complicated operations with big data sets. That means that for simple or small problems Spark may actually perform worse than some other solutions!
- Get to know the
SparkContext. - Call
print()onscto verify there's aSparkContextin your environment. print()sc.versionto see what version of Spark is running on your cluster.
# Verify SparkContext
print(sc)
# Print Spark version
print(sc.version)Spark's core data structure is the Resilient Distributed Dataset (RDD). This is a low level object that lets Spark work its magic by splitting data across multiple nodes in the cluster. However, RDDs are hard to work with directly, so in this course you'll be using the Spark DataFrame abstraction built on top of RDDs.
The Spark DataFrame was designed to behave a lot like a SQL table (a table with variables in the columns and observations in the rows). Not only are they easier to understand, DataFrames are also more optimized for complicated operations than RDDs.
When you start modifying and combining columns and rows of data, there are many ways to arrive at the same result, but some often take much longer than others. When using RDDs, it's up to the data scientist to figure out the right way to optimize the query, but the DataFrame implementation has much of this optimization built in!
To start working with Spark DataFrames, you first have to create a SparkSession object from your SparkContext. You can think of the SparkContext as your connection to the cluster and the SparkSession as your interface with that connection.
We've already created a SparkSession for you called spark, but what if you're not sure there already is one? Creating multiple SparkSessions and SparkContexts can cause issues, so it's best practice to use the SparkSession.builder.getOrCreate() method. This returns an existing SparkSession if there's already one in the environment, or creates a new one if necessary!
- Import
SparkSessionfrompyspark.sql. - Make a new
SparkSessioncalled my_spark usingSparkSession.builder.getOrCreate(). - Print
my_sparkto the console to verify it's aSparkSession.
# Import SparkSession from pyspark.sql
from pyspark.sql import SparkSession
# Create my_spark
my_spark = SparkSession.builder.getOrCreate()
# Print my_spark
print(my_spark)
#Result
<pyspark.sql.session.SparkSession object at 0x7ffaa959a490>Once you've created a SparkSession, you can start poking around to see what data is in your cluster!
Your SparkSession has an attribute called catalog which lists all the data inside the cluster. This attribute has a few methods for extracting different pieces of information.
One of the most useful is the .listTables() method, which returns the names of all the tables in your cluster as a list.
- See what tables are in your cluster by calling
spark.catalog.listTables()and printing the result!
# Print the tables in the catalog
print(spark.catalog.listTables())
#Result
[Table(name='flights', database=None, description=None, tableType='TEMPORARY', isTemporary=True)]One of the advantages of the DataFrame interface is that you can run SQL queries on the tables in your Spark cluster. If you don't have any experience with SQL, don't worry, we'll provide you with queries! (To learn more SQL, start with our Introduction to SQL course.)
As you saw in the last exercise, one of the tables in your cluster is the flights table. This table contains a row for every flight that left Portland International Airport (PDX) or Seattle-Tacoma International Airport (SEA) in 2014 and 2015.
Running a query on this table is as easy as using the .sql() method on your SparkSession. This method takes a string containing the query and returns a DataFrame with the results!
If you look closely, you'll notice that the table flights is only mentioned in the query, not as an argument to any of the methods. This is because there isn't a local object in your environment that holds that data, so it wouldn't make sense to pass the table as an argument.
Remember, we've already created a SparkSession called spark in your workspace. (It's no longer called my_spark because we created it for you!)
- Use the
.sql() methodto get the first 10 rows of theflightstable and save the result toflights10. The variablequerycontains the appropriate SQL query. - Use the DataFrame method
.show()to printflights10.
# Don't change this query
query = "FROM flights SELECT * LIMIT 10"
# Get the first 10 rows of flights
flights10 = spark.sql(query)
# Show the results
flights10.show()
#Results
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
|year|month|day|dep_time|dep_delay|arr_time|arr_delay|carrier|tailnum|flight|origin|dest|air_time|distance|hour|minute|
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
|2014| 12| 8| 658| -7| 935| -5| VX| N846VA| 1780| SEA| LAX| 132| 954| 6| 58|
|2014| 1| 22| 1040| 5| 1505| 5| AS| N559AS| 851| SEA| HNL| 360| 2677| 10| 40|
|2014| 3| 9| 1443| -2| 1652| 2| VX| N847VA| 755| SEA| SFO| 111| 679| 14| 43|
|2014| 4| 9| 1705| 45| 1839| 34| WN| N360SW| 344| PDX| SJC| 83| 569| 17| 5|
|2014| 3| 9| 754| -1| 1015| 1| AS| N612AS| 522| SEA| BUR| 127| 937| 7| 54|
|2014| 1| 15| 1037| 7| 1352| 2| WN| N646SW| 48| PDX| DEN| 121| 991| 10| 37|
|2014| 7| 2| 847| 42| 1041| 51| WN| N422WN| 1520| PDX| OAK| 90| 543| 8| 47|
|2014| 5| 12| 1655| -5| 1842| -18| VX| N361VA| 755| SEA| SFO| 98| 679| 16| 55|
|2014| 4| 19| 1236| -4| 1508| -7| AS| N309AS| 490| SEA| SAN| 135| 1050| 12| 36|
|2014| 11| 19| 1812| -3| 2352| -4| AS| N564AS| 26| SEA| ORD| 198| 1721| 18| 12|
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+Suppose you've run a query on your huge dataset and aggregated it down to something a little more manageable.
Sometimes it makes sense to then take that table and work with it locally using a tool like pandas. Spark DataFrames make that easy with the .toPandas() method. Calling this method on a Spark DataFrame returns the corresponding pandas DataFrame. It's as simple as that!
This time the query counts the number of flights to each airport from SEA and PDX.
- Run the query using the
.sql()method. Save the result inflight_counts. - Use the
.toPandas()method onflight_countsto create apandasDataFrame calledpd_counts. - Print the
.head()ofpd_countsto the console.
# Don't change this query
query = "SELECT origin, dest, COUNT(*) as N FROM flights GROUP BY origin, dest"
# Run the query
flight_counts = spark.sql(query)
# Convert the results to a pandas DataFrame
pd_counts = flight_counts.toPandas()
# Print the head of pd_counts
print(pd_counts.head())
#Results
origin dest N
0 SEA RNO 8
1 SEA DTW 98
2 SEA CLE 2
3 SEA LAX 450
4 PDX SEA 144In the last exercise, you saw how to move data from Spark to pandas. However, maybe you want to go the other direction, and put a pandas DataFrame into a Spark cluster! The SparkSession class has a method for this as well.
The .createDataFrame() method takes a pandas DataFrame and returns a Spark DataFrame.
The output of this method is stored locally, not in the SparkSession catalog. This means that you can use all the Spark DataFrame methods on it, but you can't access the data in other contexts.
For example, a SQL query (using the .sql() method) that references your DataFrame will throw an error. To access the data in this way, you have to save it as a temporary table.
You can do this using the .createTempView() Spark DataFrame method, which takes as its only argument the name of the temporary table you'd like to register. This method registers the DataFrame as a table in the catalog, but as this table is temporary, it can only be accessed from the specific SparkSession used to create the Spark DataFrame.
There is also the method .createOrReplaceTempView(). This safely creates a new temporary table if nothing was there before, or updates an existing table if one was already defined. You'll use this method to avoid running into problems with duplicate tables.
Check out the diagram to see all the different ways your Spark data structures interact with each other.
There's already a SparkSession called spark in your workspace, numpy has been imported as np, and pandas as pd.
- The code to create a
pandasDataFrame of random numbers has already been provided and saved underpd_temp. - Create a Spark DataFrame called
spark_tempby calling the Spark method.createDataFrame()withpd_tempas the argument. - Examine the list of tables in your Spark cluster and verify that the new DataFrame is not present. Remember you can use
spark.catalog.listTables()to do so. - Register the
spark_tempDataFrame you just created as a temporary table using the.createOrReplaceTempView()method. THe temporary table should be named "temp". Remember that the table name is set including it as the only argument to your method! - Examine the list of tables again.
# Create pd_temp
pd_temp = pd.DataFrame(np.random.random(10))
# Create spark_temp from pd_temp
spark_temp = spark.createDataFrame(pd_temp)
# Examine the tables in the catalog
print(spark.catalog.listTables())
# Add spark_temp to the catalog
spark_temp.createOrReplaceTempView("temp")
# Examine the tables in the catalog again
print(spark.catalog.listTables())
#Results
[Table(name='temp', database=None, description=None, tableType='TEMPORARY', isTemporary=True)]Now you know how to put data into Spark via pandas, but you're probably wondering why deal with pandas at all? Wouldn't it be easier to just read a text file straight into Spark? Of course it would!
Luckily, your SparkSession has a .read attribute which has several methods for reading different data sources into Spark DataFrames. Using these you can create a DataFrame from a .csv file just like with regular pandas DataFrames!
The variable file_path is a string with the path to the file airports.csv. This file contains information about different airports all over the world.
A SparkSession named spark is available in your workspace.
- Use the
.read.csv()method to create a Spark DataFrame calledairports- The first argument is
file_path - Pass the argument
header=Trueso that Spark knows to take the column names from the first line of the file.
- The first argument is
- Print out this DataFrame by calling
.show()
# Don't change this file path
file_path = "/usr/local/share/datasets/airports.csv"
# Read in the airports data
airports = spark.read.csv(file_path, header=True)
# Show the data
airports.show()
#Results
# Show the data
airports.show()
+---+--------------------+----------------+-----------------+----+---+---+
|faa| name| lat| lon| alt| tz|dst|
+---+--------------------+----------------+-----------------+----+---+---+
|04G| Lansdowne Airport| 41.1304722| -80.6195833|1044| -5| A|
|06A|Moton Field Munic...| 32.4605722| -85.6800278| 264| -5| A|
|06C| Schaumburg Regional| 41.9893408| -88.1012428| 801| -6| A|
|06N| Randall Airport| 41.431912| -74.3915611| 523| -5| A|
|09J|Jekyll Island Air...| 31.0744722| -81.4277778| 11| -4| A|
|0A9|Elizabethton Muni...| 36.3712222| -82.1734167|1593| -4| A|
|0G6|Williams County A...| 41.4673056| -84.5067778| 730| -5| A|
|0G7|Finger Lakes Regi...| 42.8835647| -76.7812318| 492| -5| A|
|0P2|Shoestring Aviati...| 39.7948244| -76.6471914|1000| -5| U|
|0S9|Jefferson County ...| 48.0538086| -122.8106436| 108| -8| A|
|0W3|Harford County Ai...| 39.5668378| -76.2024028| 409| -5| A|
|10C| Galt Field Airport| 42.4028889| -88.3751111| 875| -6| U|
|17G|Port Bucyrus-Craw...| 40.7815556| -82.9748056|1003| -5| A|
|19A|Jackson County Ai...| 34.1758638| -83.5615972| 951| -4| U|
|1A3|Martin Campbell F...| 35.0158056| -84.3468333|1789| -4| A|
|1B9| Mansfield Municipal| 42.0001331| -71.1967714| 122| -5| A|
|1C9|Frazier Lake Airpark|54.0133333333333|-124.768333333333| 152| -8| A|
|1CS|Clow Internationa...| 41.6959744| -88.1292306| 670| -6| U|
|1G3| Kent State Airport| 41.1513889| -81.4151111|1134| -4| A|
|1OH| Fortman Airport| 40.5553253| -84.3866186| 885| -5| U|
+---+--------------------+----------------+-----------------+----+---+---+
only showing top 20 rowsIn this chapter, you'll learn how to use the methods defined by Spark's DataFrame class to perform common data operations.
Let's look at performing column-wise operations. In Spark you can do this using the .withColumn() method, which takes two arguments. First, a string with the name of your new column, and second the new column itself.
The new column must be an object of class Column. Creating one of these is as easy as extracting a column from your DataFrame using df.colName.
Updating a Spark DataFrame is somewhat different than working in pandas because the Spark DataFrame is immutable. This means that it can't be changed, and so columns can't be updated in place.
Thus, all these methods return a new DataFrame. To overwrite the original DataFrame you must reassign the returned DataFrame using the method like so:
df = df.withColumn("newCol", df.oldCol + 1)
The above code creates a DataFrame with the same columns as df plus a new column, newCol, where every entry is equal to the corresponding entry from oldCol, plus one.
To overwrite an existing column, just pass the name of the column as the first argument!
- Use the
spark.table()method with the argument "flights" to create a DataFrame containing the values of theflightstable in the.catalog.Save it asflights. - Show the head of
flightsusingflights.show(). Check the output: the columnair_timecontains the duration of the flight in minutes. - Update
flightsto include a new column calledduration_hrs, that contains the duration of each flight in hours (you'll need to divideair_timeby the number of minutes in an hour).
# Create the DataFrame flights
flights = spark.table("flights")
# Show the head
flights.show()
# Add duration_hrs
flights = flights.withColumn("duration_hrs", flights.air_time/60)
#Results
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
|year|month|day|dep_time|dep_delay|arr_time|arr_delay|carrier|tailnum|flight|origin|dest|air_time|distance|hour|minute|
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
|2014| 12| 8| 658| -7| 935| -5| VX| N846VA| 1780| SEA| LAX| 132| 954| 6| 58|
|2014| 1| 22| 1040| 5| 1505| 5| AS| N559AS| 851| SEA| HNL| 360| 2677| 10| 40|
|2014| 3| 9| 1443| -2| 1652| 2| VX| N847VA| 755| SEA| SFO| 111| 679| 14| 43|
|2014| 4| 9| 1705| 45| 1839| 34| WN| N360SW| 344| PDX| SJC| 83| 569| 17| 5|
|2014| 3| 9| 754| -1| 1015| 1| AS| N612AS| 522| SEA| BUR| 127| 937| 7| 54|
|2014| 1| 15| 1037| 7| 1352| 2| WN| N646SW| 48| PDX| DEN| 121| 991| 10| 37|
|2014| 7| 2| 847| 42| 1041| 51| WN| N422WN| 1520| PDX| OAK| 90| 543| 8| 47|
|2014| 5| 12| 1655| -5| 1842| -18| VX| N361VA| 755| SEA| SFO| 98| 679| 16| 55|
|2014| 4| 19| 1236| -4| 1508| -7| AS| N309AS| 490| SEA| SAN| 135| 1050| 12| 36|
|2014| 11| 19| 1812| -3| 2352| -4| AS| N564AS| 26| SEA| ORD| 198| 1721| 18| 12|
|2014| 11| 8| 1653| -2| 1924| -1| AS| N323AS| 448| SEA| LAX| 130| 954| 16| 53|
|2014| 8| 3| 1120| 0| 1415| 2| AS| N305AS| 656| SEA| PHX| 154| 1107| 11| 20|
|2014| 10| 30| 811| 21| 1038| 29| AS| N433AS| 608| SEA| LAS| 127| 867| 8| 11|
|2014| 11| 12| 2346| -4| 217| -28| AS| N765AS| 121| SEA| ANC| 183| 1448| 23| 46|
|2014| 10| 31| 1314| 89| 1544| 111| AS| N713AS| 306| SEA| SFO| 129| 679| 13| 14|
|2014| 1| 29| 2009| 3| 2159| 9| UA| N27205| 1458| PDX| SFO| 90| 550| 20| 9|
|2014| 12| 17| 2015| 50| 2150| 41| AS| N626AS| 368| SEA| SMF| 76| 605| 20| 15|
|2014| 8| 11| 1017| -3| 1613| -7| WN| N8634A| 827| SEA| MDW| 216| 1733| 10| 17|
|2014| 1| 13| 2156| -9| 607| -15| AS| N597AS| 24| SEA| BOS| 290| 2496| 21| 56|
|2014| 6| 5| 1733| -12| 1945| -10| OO| N215AG| 3488| PDX| BUR| 111| 817| 17| 33|
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
only showing top 20 rowsLet's take a look at the .filter() method. As you might suspect, this is the Spark counterpart of SQL's WHERE clause. The .filter() method takes either an expression that would follow the WHERE clause of a SQL expression as a string, or a Spark Column of boolean (True/False) values.
For example, the following two expressions will produce the same output:
flights.filter("air_time > 120").show()
flights.filter(flights.air_time > 120).show()Notice that in the first case, we pass a string to .filter(). In SQL, we would write this filtering task as SELECT * FROM flights WHERE air_time > 120. Spark's .filter() can accept any expression that could go in the WHERE clause of a SQL query (in this case, "air_time > 120"), as long as it is passed as a string. Notice that in this case, we do not reference the name of the table in the string -- as we wouldn't in the SQL request.
In the second case, we actually pass a column of boolean values to .filter(). Remember that flights.air_time > 120 returns a column of boolean values that has True in place of those records in flights.air_time that are over 120, and False otherwise.
- Use the
.filter()method to find all the flights that flew over 1000 miles two ways:- First, pass a SQL string to
.filter()that checks whether the distance is greater than 1000. Save this aslong_flights1. - Then pass a column of boolean values to
.filter()that checks the same thing. Save this aslong_flights2.
- First, pass a SQL string to
- Use
.show()to print heads of both DataFrames and make sure they're actually equal!
# Filter flights by passing a string
long_flights1 = flights.filter("distance > 1000")
# Filter flights by passing a column of boolean values
long_flights2 = flights.filter(flights.distance > 1000)
# Print the data to check they're equal
long_flights1.show()
long_flights2.show()
#Results
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
|year|month|day|dep_time|dep_delay|arr_time|arr_delay|carrier|tailnum|flight|origin|dest|air_time|distance|hour|minute|
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
|2014| 1| 22| 1040| 5| 1505| 5| AS| N559AS| 851| SEA| HNL| 360| 2677| 10| 40|
|2014| 4| 19| 1236| -4| 1508| -7| AS| N309AS| 490| SEA| SAN| 135| 1050| 12| 36|
|2014| 11| 19| 1812| -3| 2352| -4| AS| N564AS| 26| SEA| ORD| 198| 1721| 18| 12|
|2014| 8| 3| 1120| 0| 1415| 2| AS| N305AS| 656| SEA| PHX| 154| 1107| 11| 20|
|2014| 11| 12| 2346| -4| 217| -28| AS| N765AS| 121| SEA| ANC| 183| 1448| 23| 46|
|2014| 8| 11| 1017| -3| 1613| -7| WN| N8634A| 827| SEA| MDW| 216| 1733| 10| 17|
|2014| 1| 13| 2156| -9| 607| -15| AS| N597AS| 24| SEA| BOS| 290| 2496| 21| 56|
|2014| 9| 26| 610| -5| 1523| 65| US| N127UW| 616| SEA| PHL| 293| 2378| 6| 10|
|2014| 12| 4| 954| -6| 1348| -17| HA| N395HA| 29| SEA| OGG| 333| 2640| 9| 54|
|2014| 6| 4| 1115| 0| 1346| -3| AS| N461AS| 488| SEA| SAN| 133| 1050| 11| 15|
|2014| 6| 26| 2054| -1| 2318| -6| B6| N590JB| 907| SEA| ANC| 179| 1448| 20| 54|
|2014| 6| 7| 1823| -7| 2112| -28| AS| N512AS| 815| SEA| LIH| 335| 2701| 18| 23|
|2014| 4| 30| 801| 1| 1757| 90| AS| N407AS| 18| SEA| MCO| 342| 2554| 8| 1|
|2014| 11| 29| 905| 155| 1655| 170| DL| N824DN| 1598| SEA| ATL| 229| 2182| 9| 5|
|2014| 6| 2| 2222| 7| 55| 15| AS| N402AS| 99| SEA| ANC| 190| 1448| 22| 22|
|2014| 11| 15| 1034| -6| 1414| -26| AS| N589AS| 794| SEA| ABQ| 139| 1180| 10| 34|
|2014| 10| 20| 1328| -1| 1949| 4| UA| N68805| 1212| SEA| IAH| 228| 1874| 13| 28|
|2014| 12| 16| 1500| 0| 1906| 19| US| N662AW| 500| SEA| PHX| 151| 1107| 15| 0|
|2014| 11| 19| 1319| -6| 1821| -14| DL| N309US| 2164| PDX| MSP| 169| 1426| 13| 19|
|2014| 5| 21| 515| 0| 757| 0| US| N172US| 593| SEA| PHX| 143| 1107| 5| 15|
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
only showing top 20 rows
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
|year|month|day|dep_time|dep_delay|arr_time|arr_delay|carrier|tailnum|flight|origin|dest|air_time|distance|hour|minute|
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
|2014| 1| 22| 1040| 5| 1505| 5| AS| N559AS| 851| SEA| HNL| 360| 2677| 10| 40|
|2014| 4| 19| 1236| -4| 1508| -7| AS| N309AS| 490| SEA| SAN| 135| 1050| 12| 36|
|2014| 11| 19| 1812| -3| 2352| -4| AS| N564AS| 26| SEA| ORD| 198| 1721| 18| 12|
|2014| 8| 3| 1120| 0| 1415| 2| AS| N305AS| 656| SEA| PHX| 154| 1107| 11| 20|
|2014| 11| 12| 2346| -4| 217| -28| AS| N765AS| 121| SEA| ANC| 183| 1448| 23| 46|
|2014| 8| 11| 1017| -3| 1613| -7| WN| N8634A| 827| SEA| MDW| 216| 1733| 10| 17|
|2014| 1| 13| 2156| -9| 607| -15| AS| N597AS| 24| SEA| BOS| 290| 2496| 21| 56|
|2014| 9| 26| 610| -5| 1523| 65| US| N127UW| 616| SEA| PHL| 293| 2378| 6| 10|
|2014| 12| 4| 954| -6| 1348| -17| HA| N395HA| 29| SEA| OGG| 333| 2640| 9| 54|
|2014| 6| 4| 1115| 0| 1346| -3| AS| N461AS| 488| SEA| SAN| 133| 1050| 11| 15|
|2014| 6| 26| 2054| -1| 2318| -6| B6| N590JB| 907| SEA| ANC| 179| 1448| 20| 54|
|2014| 6| 7| 1823| -7| 2112| -28| AS| N512AS| 815| SEA| LIH| 335| 2701| 18| 23|
|2014| 4| 30| 801| 1| 1757| 90| AS| N407AS| 18| SEA| MCO| 342| 2554| 8| 1|
|2014| 11| 29| 905| 155| 1655| 170| DL| N824DN| 1598| SEA| ATL| 229| 2182| 9| 5|
|2014| 6| 2| 2222| 7| 55| 15| AS| N402AS| 99| SEA| ANC| 190| 1448| 22| 22|
|2014| 11| 15| 1034| -6| 1414| -26| AS| N589AS| 794| SEA| ABQ| 139| 1180| 10| 34|
|2014| 10| 20| 1328| -1| 1949| 4| UA| N68805| 1212| SEA| IAH| 228| 1874| 13| 28|
|2014| 12| 16| 1500| 0| 1906| 19| US| N662AW| 500| SEA| PHX| 151| 1107| 15| 0|
|2014| 11| 19| 1319| -6| 1821| -14| DL| N309US| 2164| PDX| MSP| 169| 1426| 13| 19|
|2014| 5| 21| 515| 0| 757| 0| US| N172US| 593| SEA| PHX| 143| 1107| 5| 15|
+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+----+--------+--------+----+------+
only showing top 20 rowsThe Spark variant of SQL's SELECT is the .select() method. This method takes multiple arguments - one for each column you want to select. These arguments can either be the column name as a string (one for each column) or a column object (using the df.colName syntax). When you pass a column object, you can perform operations like addition or subtraction on the column to change the data contained in it, much like inside .withColumn().
The difference between .select() and .withColumn() methods is that .select() returns only the columns you specify, while .withColumn() returns all the columns of the DataFrame in addition to the one you defined. It's often a good idea to drop columns you don't need at the beginning of an operation so that you're not dragging around extra data as you're wrangling. In this case, you would use .select() and not .withColumn().
- Select the columns "tailnum", "origin", and "dest" from
flightsby passing the column names as strings. Save this asselected1. - Select the columns "origin", "dest", and "carrier" using the
df.colNamesyntax and then filter the result using both of the filters already defined for you (filterA and filterB) to only keep flights from SEA to PDX. Save this asselected2.
# Select the first set of columns
selected1 = flights.select("tailnum", "origin", "dest")
# Select the second set of columns
temp = flights.select(flights.origin, flights.dest, flights.carrier)
# Define first filter
filterA = flights.origin == "SEA"
# Define second filter
filterB = flights.dest == "PDX"
# Filter the data, first by filterA then by filterB
selected2 = temp.filter(filterA).filter(filterB)Similar to SQL, you can also use the .select() method to perform column-wise operations. When you're selecting a column using the df.colName notation, you can perform any column operation and the .select() method will return the transformed column. For example,
flights.select(flights.air_time/60)returns a column of flight durations in hours instead of minutes. You can also use the .alias() method to rename a column you're selecting. So if you wanted to .select() the column duration_hrs (which isn't in your DataFrame) you could do
flights.select((flights.air_time/60).alias("duration_hrs"))The equivalent Spark DataFrame method .selectExpr() takes SQL expressions as a string:
flights.selectExpr("air_time/60 as duration_hrs")with the SQL as keyword being equivalent to the .alias() method. To select multiple columns, you can pass multiple strings.
Create a table of the average speed of each flight both ways.
- Calculate average speed by dividing the distance by the
air_time(converted to hours). Use the.alias()method name this column "avg_speed". Save the output as the variableavg_speed. - Select the columns "origin", "dest", "tailnum", and
avg_speed(without quotes!). Save this asspeed1. - Create the same table using
.selectExpr()and a string containing a SQL expression. Save this asspeed2.
# Define avg_speed
avg_speed = (flights.distance/(flights.air_time/60)).alias("avg_speed")
# Select the correct columns
speed1 = flights.select("origin", "dest", "tailnum", avg_speed)
# Create the same table using a SQL expression
speed2 = flights.selectExpr("origin", "dest", "tailnum", "distance/(air_time/60) as avg_speed")All of the common aggregation methods, like .min(), .max(), and .count() are GroupedData methods. These are created by calling the .groupBy() DataFrame method. You'll learn exactly what that means in a few exercises. For now, all you have to do to use these functions is call that method on your DataFrame. For example, to find the minimum value of a column, col, in a DataFrame, df, you could do
df.groupBy().min("col").show()This creates a GroupedData object (so you can use the .min() method), then finds the minimum value in col, and returns it as a DataFrame.
Now you're ready to do some aggregating of your own!
- Find the length of the shortest (in terms of distance) flight that left PDX by first
.filter()ing and using the.min()method. Perform the filtering by referencing the column directly, not passing a SQL string. - Find the length of the longest (in terms of time) flight that left SEA by
filter()ing and using the.max()method. Perform the filtering by referencing the column directly, not passing a SQL string.
# Find the shortest flight from PDX in terms of distance
flights.filter(flights.origin == "PDX").groupBy().min("distance").show()
# Find the longest flight from SEA in terms of air time
flights.filter(flights.origin == "SEA").groupBy().max("air_time").show()
# Results
+-------------+
|min(distance)|
+-------------+
| 106|
+-------------+
+-------------+
|max(air_time)|
+-------------+
| 409|
+-------------+
#The filter() method is used here, and the condition is specified by directly referencing the origin column, which matches the requirement to not use a SQL string.To get you familiar with more of the built in aggregation methods, here's a few more exercises involving the flights table!
- Use the
.avg()method to get the average air time of Delta Airlines flights (where the carrier column has the value "DL") that left SEA. The place of departure is stored in the column origin.show()the result. - Use the
.sum()method to get the total number of hours all planes in this dataset spent in the air by creating a column called duration_hrs from the column air_time.show()the result.
# Average duration of Delta flights
flights.filter(flights.origin == "SEA").filter(flights.carrier == "DL").groupBy().avg("air_time").show()
# Total hours in the air
flights.withColumn("duration_hrs", flights.air_time/60).groupBy().sum("duration_hrs").show()
#Result
+------------------+
| avg(air_time)|
+------------------+
|188.20689655172413|
+------------------+
+------------------+
| sum(duration_hrs)|
+------------------+
|25289.600000000126|
+------------------+Part of what makes aggregating so powerful is the addition of groups. PySpark has a whole class devoted to grouped data frames: pyspark.sql.GroupedData, which you saw in the last two exercises.
You've learned how to create a grouped DataFrame by calling the .groupBy() method on a DataFrame with no arguments.
Now you'll see that when you pass the name of one or more columns in your DataFrame to the .groupBy() method, the aggregation methods behave like when you use a GROUP BY statement in a SQL query!
- Create a DataFrame called
by_planethat is grouped by the column tailnum. - Use the
.count()method with no arguments to count the number of flights each plane made. - Create a DataFrame called
by_originthat is grouped by the column origin. - Find the
.avg()of theair_timecolumn to find average duration of flights from PDX and SEA.
# Group by tailnum
by_plane = flights.groupBy("tailnum")
# Number of flights each plane made
by_plane.count().show()
# Group by origin
by_origin = flights.groupBy("origin")
# Average duration of flights from PDX and SEA
by_origin.avg("air_time").show()
#Result
+-------+-----+
|tailnum|count|
+-------+-----+
| N442AS| 38|
| N102UW| 2|
| N36472| 4|
| N38451| 4|
| N73283| 4|
| N513UA| 2|
| N954WN| 5|
| N388DA| 3|
| N567AA| 1|
| N516UA| 2|
| N927DN| 1|
| N8322X| 1|
| N466SW| 1|
| N6700| 1|
| N607AS| 45|
| N622SW| 4|
| N584AS| 31|
| N914WN| 4|
| N654AW| 2|
| N336NW| 1|
+-------+-----+
only showing top 20 rows
+------+------------------+
|origin| avg(air_time)|
+------+------------------+
| SEA| 160.4361496051259|
| PDX|137.11543248288737|
+------+------------------+In addition to the GroupedData methods you've already seen, there is also the .agg() method. This method lets you pass an aggregate column expression that uses any of the aggregate functions from the pyspark.sql.functions submodule.
This submodule contains many useful functions for computing things like standard deviations. All the aggregation functions in this submodule take the name of a column in a GroupedData table.
- Import the submodule
pyspark.sql.functionsasF. - Create a
GroupedDatatable calledby_month_destthat's grouped by both the month and dest columns. Refer to the two columns by passing both strings as separate arguments. - Use the
.avg()method on theby_month_destDataFrame to get theaverage dep_delayin each month for each destination. - Find the standard deviation of
dep_delayby using the.agg()method with the functionF.stddev().
# Import pyspark.sql.functions as F
import pyspark.sql.functions as F
# Group by month and dest
by_month_dest = flights.groupBy("month", "dest")
# Average departure delay by month and destination
by_month_dest.avg("dep_delay").show()
# Standard deviation of departure delay
by_month_dest.agg(F.stddev("dep_delay")).show()
#Result
+-----+----+--------------------+
|month|dest| avg(dep_delay)|
+-----+----+--------------------+
| 11| TUS| -2.3333333333333335|
| 11| ANC| 7.529411764705882|
| 1| BUR| -1.45|
| 1| PDX| -5.6923076923076925|
| 6| SBA| -2.5|
| 5| LAX|-0.15789473684210525|
| 10| DTW| 2.6|
| 6| SIT| -1.0|
| 10| DFW| 18.176470588235293|
| 3| FAI| -2.2|
| 10| SEA| -0.8|
| 2| TUS| -0.6666666666666666|
| 12| OGG| 25.181818181818183|
| 9| DFW| 4.066666666666666|
| 5| EWR| 14.25|
| 3| RDM| -6.2|
| 8| DCA| 2.6|
| 7| ATL| 4.675675675675675|
| 4| JFK| 0.07142857142857142|
| 10| SNA| -1.1333333333333333|
+-----+----+--------------------+
only showing top 20 rows
+-----+----+----------------------+
|month|dest|stddev_samp(dep_delay)|
+-----+----+----------------------+
| 11| TUS| 3.0550504633038935|
| 11| ANC| 18.604716401245316|
| 1| BUR| 15.22627576540667|
| 1| PDX| 5.677214918493858|
| 6| SBA| 2.380476142847617|
| 5| LAX| 13.36268698685904|
| 10| DTW| 5.639148871948674|
| 6| SIT| null|
| 10| DFW| 45.53019017606675|
| 3| FAI| 3.1144823004794873|
| 10| SEA| 18.70523227029577|
| 2| TUS| 14.468356276140469|
| 12| OGG| 82.64480404939947|
| 9| DFW| 21.728629347782924|
| 5| EWR| 42.41595968929191|
| 3| RDM| 2.16794833886788|
| 8| DCA| 9.946523680831074|
| 7| ATL| 22.767001039582183|
| 4| JFK| 8.156774303176903|
| 10| SNA| 13.726234873756304|
+-----+----+----------------------+
only showing top 20 rowsAnother very common data operation is the join. Joins are a whole topic unto themselves, so in this course we'll just look at simple joins. If you'd like to learn more about joins, you can take a look here.
A join will combine two different tables along a column that they share. This column is called the key. Examples of keys here include the tailnum and carrier columns from the flights table.
For example, suppose that you want to know more information about the plane that flew a flight than just the tail number. This information isn't in the flights table because the same plane flies many different flights over the course of two years, so including this information in every row would result in a lot of duplication. To avoid this, you'd have a second table that has only one row for each plane and whose columns list all the information about the plane, including its tail number. You could call this table planes
When you join the flights table to this table of airplane information, you're adding all the columns from the planes table to the flights table. To fill these columns with information, you'll look at the tail number from the flights table and find the matching one in the planes table, and then use that row to fill out all the new columns.
In PySpark, joins are performed using the DataFrame method .join(). This method takes three arguments. The first is the second DataFrame that you want to join with the first one. The second argument, on, is the name of the key column(s) as a string. The names of the key column(s) must be the same in each table. The third argument, how, specifies the kind of join to perform. In this course we'll always use the value how="leftouter".
- Examine the
airportsDataFrame by calling.show(). Note which key column will let you joinairportsto theflightstable. - Rename the
faacolumn inairportstodestby re-assigning the result ofairports.withColumnRenamed("faa", "dest")toairports. - Join the
flightswith theairportsDataFrame on thedestcolumn by calling the.join()method onflights. Save the result asflights_with_airports.- The first argument should be the other DataFrame,
airports. - The argument
onshould be the key column. - The argument
howshould be"leftouter".
- The first argument should be the other DataFrame,
- Call
.show()onflights_with_airportsto examine the data again. Note the new information that has been added.
# Examine the data
print(airports.show())
#Result
+---+--------------------+----------------+-----------------+----+---+---+
|faa| name| lat| lon| alt| tz|dst|
+---+--------------------+----------------+-----------------+----+---+---+
|04G| Lansdowne Airport| 41.1304722| -80.6195833|1044| -5| A|
|06A|Moton Field Munic...| 32.4605722| -85.6800278| 264| -5| A|
|06C| Schaumburg Regional| 41.9893408| -88.1012428| 801| -6| A|
|06N| Randall Airport| 41.431912| -74.3915611| 523| -5| A|
|09J|Jekyll Island Air...| 31.0744722| -81.4277778| 11| -4| A|
|0A9|Elizabethton Muni...| 36.3712222| -82.1734167|1593| -4| A|
|0G6|Williams County A...| 41.4673056| -84.5067778| 730| -5| A|
|0G7|Finger Lakes Regi...| 42.8835647| -76.7812318| 492| -5| A|
|0P2|Shoestring Aviati...| 39.7948244| -76.6471914|1000| -5| U|
|0S9|Jefferson County ...| 48.0538086| -122.8106436| 108| -8| A|
|0W3|Harford County Ai...| 39.5668378| -76.2024028| 409| -5| A|
|10C| Galt Field Airport| 42.4028889| -88.3751111| 875| -6| U|
|17G|Port Bucyrus-Craw...| 40.7815556| -82.9748056|1003| -5| A|
|19A|Jackson County Ai...| 34.1758638| -83.5615972| 951| -4| U|
|1A3|Martin Campbell F...| 35.0158056| -84.3468333|1789| -4| A|
|1B9| Mansfield Municipal| 42.0001331| -71.1967714| 122| -5| A|
|1C9|Frazier Lake Airpark|54.0133333333333|-124.768333333333| 152| -8| A|
|1CS|Clow Internationa...| 41.6959744| -88.1292306| 670| -6| U|
|1G3| Kent State Airport| 41.1513889| -81.4151111|1134| -4| A|
|1OH| Fortman Airport| 40.5553253| -84.3866186| 885| -5| U|
+---+--------------------+----------------+-----------------+----+---+---+
only showing top 20 rows
# Rename the faa column
airports = airports.withColumnRenamed("faa", "dest")
# Examine the data
print(airports.show())
#Result
+----+--------------------+----------------+-----------------+----+---+---+
|dest| name| lat| lon| alt| tz|dst|
+----+--------------------+----------------+-----------------+----+---+---+
| 04G| Lansdowne Airport| 41.1304722| -80.6195833|1044| -5| A|
| 06A|Moton Field Munic...| 32.4605722| -85.6800278| 264| -5| A|
| 06C| Schaumburg Regional| 41.9893408| -88.1012428| 801| -6| A|
| 06N| Randall Airport| 41.431912| -74.3915611| 523| -5| A|
| 09J|Jekyll Island Air...| 31.0744722| -81.4277778| 11| -4| A|
| 0A9|Elizabethton Muni...| 36.3712222| -82.1734167|1593| -4| A|
| 0G6|Williams County A...| 41.4673056| -84.5067778| 730| -5| A|
| 0G7|Finger Lakes Regi...| 42.8835647| -76.7812318| 492| -5| A|
| 0P2|Shoestring Aviati...| 39.7948244| -76.6471914|1000| -5| U|
| 0S9|Jefferson County ...| 48.0538086| -122.8106436| 108| -8| A|
| 0W3|Harford County Ai...| 39.5668378| -76.2024028| 409| -5| A|
| 10C| Galt Field Airport| 42.4028889| -88.3751111| 875| -6| U|
| 17G|Port Bucyrus-Craw...| 40.7815556| -82.9748056|1003| -5| A|
| 19A|Jackson County Ai...| 34.1758638| -83.5615972| 951| -4| U|
| 1A3|Martin Campbell F...| 35.0158056| -84.3468333|1789| -4| A|
| 1B9| Mansfield Municipal| 42.0001331| -71.1967714| 122| -5| A|
| 1C9|Frazier Lake Airpark|54.0133333333333|-124.768333333333| 152| -8| A|
| 1CS|Clow Internationa...| 41.6959744| -88.1292306| 670| -6| U|
| 1G3| Kent State Airport| 41.1513889| -81.4151111|1134| -4| A|
| 1OH| Fortman Airport| 40.5553253| -84.3866186| 885| -5| U|
+----+--------------------+----------------+-----------------+----+---+---+
only showing top 20 rows
# Join the DataFrames
flights_with_airports = flights.join(airports, "dest", "leftouter")
# Examine the new DataFrame
print(flights_with_airports.show())
# Result
+----+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+--------+--------+----+------+--------------------+---------+-----------+----+---+---+
|dest|year|month|day|dep_time|dep_delay|arr_time|arr_delay|carrier|tailnum|flight|origin|air_time|distance|hour|minute| name| lat| lon| alt| tz|dst|
+----+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+--------+--------+----+------+--------------------+---------+-----------+----+---+---+
| LAX|2014| 12| 8| 658| -7| 935| -5| VX| N846VA| 1780| SEA| 132| 954| 6| 58| Los Angeles Intl|33.942536|-118.408075| 126| -8| A|
| HNL|2014| 1| 22| 1040| 5| 1505| 5| AS| N559AS| 851| SEA| 360| 2677| 10| 40| Honolulu Intl|21.318681|-157.922428| 13|-10| N|
| SFO|2014| 3| 9| 1443| -2| 1652| 2| VX| N847VA| 755| SEA| 111| 679| 14| 43| San Francisco Intl|37.618972|-122.374889| 13| -8| A|
| SJC|2014| 4| 9| 1705| 45| 1839| 34| WN| N360SW| 344| PDX| 83| 569| 17| 5|Norman Y Mineta S...| 37.3626|-121.929022| 62| -8| A|
| BUR|2014| 3| 9| 754| -1| 1015| 1| AS| N612AS| 522| SEA| 127| 937| 7| 54| Bob Hope|34.200667|-118.358667| 778| -8| A|
| DEN|2014| 1| 15| 1037| 7| 1352| 2| WN| N646SW| 48| PDX| 121| 991| 10| 37| Denver Intl|39.861656|-104.673178|5431| -7| A|
| OAK|2014| 7| 2| 847| 42| 1041| 51| WN| N422WN| 1520| PDX| 90| 543| 8| 47|Metropolitan Oakl...|37.721278|-122.220722| 9| -8| A|
| SFO|2014| 5| 12| 1655| -5| 1842| -18| VX| N361VA| 755| SEA| 98| 679| 16| 55| San Francisco Intl|37.618972|-122.374889| 13| -8| A|
| SAN|2014| 4| 19| 1236| -4| 1508| -7| AS| N309AS| 490| SEA| 135| 1050| 12| 36| San Diego Intl|32.733556|-117.189667| 17| -8| A|
| ORD|2014| 11| 19| 1812| -3| 2352| -4| AS| N564AS| 26| SEA| 198| 1721| 18| 12| Chicago Ohare Intl|41.978603| -87.904842| 668| -6| A|
| LAX|2014| 11| 8| 1653| -2| 1924| -1| AS| N323AS| 448| SEA| 130| 954| 16| 53| Los Angeles Intl|33.942536|-118.408075| 126| -8| A|
| PHX|2014| 8| 3| 1120| 0| 1415| 2| AS| N305AS| 656| SEA| 154| 1107| 11| 20|Phoenix Sky Harbo...|33.434278|-112.011583|1135| -7| N|
| LAS|2014| 10| 30| 811| 21| 1038| 29| AS| N433AS| 608| SEA| 127| 867| 8| 11| Mc Carran Intl|36.080056| -115.15225|2141| -8| A|
| ANC|2014| 11| 12| 2346| -4| 217| -28| AS| N765AS| 121| SEA| 183| 1448| 23| 46|Ted Stevens Ancho...|61.174361|-149.996361| 152| -9| A|
| SFO|2014| 10| 31| 1314| 89| 1544| 111| AS| N713AS| 306| SEA| 129| 679| 13| 14| San Francisco Intl|37.618972|-122.374889| 13| -8| A|
| SFO|2014| 1| 29| 2009| 3| 2159| 9| UA| N27205| 1458| PDX| 90| 550| 20| 9| San Francisco Intl|37.618972|-122.374889| 13| -8| A|
| SMF|2014| 12| 17| 2015| 50| 2150| 41| AS| N626AS| 368| SEA| 76| 605| 20| 15| Sacramento Intl|38.695417|-121.590778| 27| -8| A|
| MDW|2014| 8| 11| 1017| -3| 1613| -7| WN| N8634A| 827| SEA| 216| 1733| 10| 17| Chicago Midway Intl|41.785972| -87.752417| 620| -6| A|
| BOS|2014| 1| 13| 2156| -9| 607| -15| AS| N597AS| 24| SEA| 290| 2496| 21| 56|General Edward La...|42.364347| -71.005181| 19| -5| A|
| BUR|2014| 6| 5| 1733| -12| 1945| -10| OO| N215AG| 3488| PDX| 111| 817| 17| 33| Bob Hope|34.200667|-118.358667| 778| -8| A|
+----+----+-----+---+--------+---------+--------+---------+-------+-------+------+------+--------+--------+----+------+--------------------+---------+-----------+----+---+---+
only showing top 20 rowsPySpark has built-in, cutting-edge machine learning routines, along with utilities to create full machine learning pipelines.
At the core of the pyspark.ml module are the Transformer and Estimator classes. Almost every other class in the module behaves similarly to these two basic classes.
Transformer classes have a .transform() method that takes a DataFrame and returns a new DataFrame; usually the original one with a new column appended. For example, you might use the class Bucketizer to create discrete bins from a continuous feature or the class PCA to reduce the dimensionality of your dataset using principal component analysis.
Estimator classes all implement a .fit() method. These methods also take a DataFrame, but instead of returning another DataFrame they return a model object. This can be something like a StringIndexerModel for including categorical data saved as strings in your models, or a RandomForestModel that uses the random forest algorithm for classification or regression.
In the next two chapters you'll be working to build a model that predicts whether or not a flight will be delayed based on the flights data we've been working with. This model will also include information about the plane that flew the route, so the first step is to join the two tables: flights and planes!
- First, rename the
yearcolumn ofplanestoplane_yearto avoid duplicate column names. - Create a new DataFrame called
model_databy joining theflightstable withplanesusing thetailnumcolumn as the key.
# Rename year column
planes = planes.withColumnRenamed("year", "plane_year")
print(planes.show())
#Result
+-------+----------+--------------------+----------------+--------+-------+-----+-----+---------+
|tailnum|plane_year| type| manufacturer| model|engines|seats|speed| engine|
+-------+----------+--------------------+----------------+--------+-------+-----+-----+---------+
| N102UW| 1998|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N103US| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N104UW| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N105UW| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N107US| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N108UW| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N109UW| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N110UW| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N111US| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N11206| 2000|Fixed wing multi ...| BOEING| 737-824| 2| 149| NA|Turbo-fan|
| N112US| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N113UW| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N114UW| 1999|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N117UW| 2000|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N118US| 2000|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N119US| 2000|Fixed wing multi ...|AIRBUS INDUSTRIE|A320-214| 2| 182| NA|Turbo-fan|
| N1200K| 1998|Fixed wing multi ...| BOEING| 767-332| 2| 330| NA|Turbo-fan|
| N1201P| 1998|Fixed wing multi ...| BOEING| 767-332| 2| 330| NA|Turbo-fan|
| N12114| 1995|Fixed wing multi ...| BOEING| 757-224| 2| 178| NA|Turbo-jet|
| N121DE| 1987|Fixed wing multi ...| BOEING| 767-332| 2| 330| NA|Turbo-fan|
+-------+----------+--------------------+----------------+--------+-------+-----+-----+---------+
only showing top 20 rows
# Join the DataFrames
model_data = flights.join(planes, on="tailnum", how="leftouter")
print(model_data.show())
#Result
+-------+----+-----+---+--------+---------+--------+---------+-------+------+------+----+--------+--------+----+------+----------+--------------------+--------------+-----------+-------+-----+-----+---------+
|tailnum|year|month|day|dep_time|dep_delay|arr_time|arr_delay|carrier|flight|origin|dest|air_time|distance|hour|minute|plane_year| type| manufacturer| model|engines|seats|speed| engine|
+-------+----+-----+---+--------+---------+--------+---------+-------+------+------+----+--------+--------+----+------+----------+--------------------+--------------+-----------+-------+-----+-----+---------+
| N846VA|2014| 12| 8| 658| -7| 935| -5| VX| 1780| SEA| LAX| 132| 954| 6| 58| 2011|Fixed wing multi ...| AIRBUS| A320-214| 2| 182| NA|Turbo-fan|
| N559AS|2014| 1| 22| 1040| 5| 1505| 5| AS| 851| SEA| HNL| 360| 2677| 10| 40| 2006|Fixed wing multi ...| BOEING| 737-890| 2| 149| NA|Turbo-fan|
| N847VA|2014| 3| 9| 1443| -2| 1652| 2| VX| 755| SEA| SFO| 111| 679| 14| 43| 2011|Fixed wing multi ...| AIRBUS| A320-214| 2| 182| NA|Turbo-fan|
| N360SW|2014| 4| 9| 1705| 45| 1839| 34| WN| 344| PDX| SJC| 83| 569| 17| 5| 1992|Fixed wing multi ...| BOEING| 737-3H4| 2| 149| NA|Turbo-fan|
| N612AS|2014| 3| 9| 754| -1| 1015| 1| AS| 522| SEA| BUR| 127| 937| 7| 54| 1999|Fixed wing multi ...| BOEING| 737-790| 2| 151| NA|Turbo-jet|
| N646SW|2014| 1| 15| 1037| 7| 1352| 2| WN| 48| PDX| DEN| 121| 991| 10| 37| 1997|Fixed wing multi ...| BOEING| 737-3H4| 2| 149| NA|Turbo-fan|
| N422WN|2014| 7| 2| 847| 42| 1041| 51| WN| 1520| PDX| OAK| 90| 543| 8| 47| 2002|Fixed wing multi ...| BOEING| 737-7H4| 2| 140| NA|Turbo-fan|
| N361VA|2014| 5| 12| 1655| -5| 1842| -18| VX| 755| SEA| SFO| 98| 679| 16| 55| 2013|Fixed wing multi ...| AIRBUS| A320-214| 2| 182| NA|Turbo-fan|
| N309AS|2014| 4| 19| 1236| -4| 1508| -7| AS| 490| SEA| SAN| 135| 1050| 12| 36| 2001|Fixed wing multi ...| BOEING| 737-990| 2| 149| NA|Turbo-jet|
| N564AS|2014| 11| 19| 1812| -3| 2352| -4| AS| 26| SEA| ORD| 198| 1721| 18| 12| 2006|Fixed wing multi ...| BOEING| 737-890| 2| 149| NA|Turbo-fan|
| N323AS|2014| 11| 8| 1653| -2| 1924| -1| AS| 448| SEA| LAX| 130| 954| 16| 53| 2004|Fixed wing multi ...| BOEING| 737-990| 2| 149| NA|Turbo-jet|
| N305AS|2014| 8| 3| 1120| 0| 1415| 2| AS| 656| SEA| PHX| 154| 1107| 11| 20| 2001|Fixed wing multi ...| BOEING| 737-990| 2| 149| NA|Turbo-jet|
| N433AS|2014| 10| 30| 811| 21| 1038| 29| AS| 608| SEA| LAS| 127| 867| 8| 11| 2013|Fixed wing multi ...| BOEING| 737-990ER| 2| 222| NA|Turbo-fan|
| N765AS|2014| 11| 12| 2346| -4| 217| -28| AS| 121| SEA| ANC| 183| 1448| 23| 46| 1992|Fixed wing multi ...| BOEING| 737-4Q8| 2| 149| NA|Turbo-fan|
| N713AS|2014| 10| 31| 1314| 89| 1544| 111| AS| 306| SEA| SFO| 129| 679| 13| 14| 1999|Fixed wing multi ...| BOEING| 737-490| 2| 149| NA|Turbo-jet|
| N27205|2014| 1| 29| 2009| 3| 2159| 9| UA| 1458| PDX| SFO| 90| 550| 20| 9| 2000|Fixed wing multi ...| BOEING| 737-824| 2| 149| NA|Turbo-fan|
| N626AS|2014| 12| 17| 2015| 50| 2150| 41| AS| 368| SEA| SMF| 76| 605| 20| 15| 2001|Fixed wing multi ...| BOEING| 737-790| 2| 151| NA|Turbo-jet|
| N8634A|2014| 8| 11| 1017| -3| 1613| -7| WN| 827| SEA| MDW| 216| 1733| 10| 17| 2014|Fixed wing multi ...| BOEING| 737-8H4| 2| 140| NA|Turbo-fan|
| N597AS|2014| 1| 13| 2156| -9| 607| -15| AS| 24| SEA| BOS| 290| 2496| 21| 56| 2008|Fixed wing multi ...| BOEING| 737-890| 2| 149| NA|Turbo-fan|
| N215AG|2014| 6| 5| 1733| -12| 1945| -10| OO| 3488| PDX| BUR| 111| 817| 17| 33| 2001|Fixed wing multi ...|BOMBARDIER INC|CL-600-2C10| 2| 80| NA|Turbo-fan|
+-------+----+-----+---+--------+---------+--------+---------+-------+------+------+----+--------+--------+----+------+----------+--------------------+--------------+-----------+-------+-----+-----+---------+
only showing top 20 rowsGood work! Before you get started modeling, it's important to know that Spark only handles numeric data. That means all of the columns in your DataFrame must be either integers or decimals (called 'doubles' in Spark).
When we imported our data, we let Spark guess what kind of information each column held. Unfortunately, Spark doesn't always guess right and you can see that some of the columns in our DataFrame are strings containing numbers as opposed to actual numeric values.
To remedy this, you can use the .cast() method in combination with the .withColumn() method. It's important to note that .cast() works on columns, while .withColumn() works on DataFrames.
The only argument you need to pass to .cast() is the kind of value you want to create, in string form. For example, to create integers, you'll pass the argument "integer" and for decimal numbers you'll use "double".
You can put this call to .cast() inside a call to .withColumn() to overwrite the already existing column, just like you did in the previous chapter!
Now you'll use the .cast() method you learned in the previous exercise to convert all the appropriate columns from your DataFrame model_data to integers!
To convert the type of a column using the .cast() method, you can write code like this:
dataframe = dataframe.withColumn("col", dataframe.col.cast("new_type"))-
Use the method
.withColumn()to.cast()the following columns to type "integer". Access the columns using thedf.colnotation:- model_data.arr_delay
- model_data.air_time
- model_data.month
- model_data.plane_year
# Cast the columns to integers
model_data = model_data.withColumn("arr_delay", model_data.arr_delay.cast("integer"))
model_data = model_data.withColumn("air_time", model_data.air_time.cast("integer"))
model_data = model_data.withColumn("month", model_data.month.cast("integer"))
model_data = model_data.withColumn("plane_year", model_data.plane_year.cast("integer"))In the last exercise, you converted the column plane_year to an integer. This column holds the year each plane was manufactured. However, your model will use the planes' age, which is slightly different from the year it was made!
- Create the column
plane_ageusing the.withColumn()method and subtracting the year of manufacture (columnplane_year) from the year (columnyear) of the flight.
# Create the column plane_age
model_data = model_data.withColumn("plane_age", model_data.year - model_data.plane_year)Consider that you're modeling a yes or no question: is the flight late? However, your data contains the arrival delay in minutes for each flight. Thus, you'll need to create a boolean column which indicates whether the flight was late or not!
- Use the
.withColumn()method to create the columnis_late. This column is equal tomodel_data.arr_delay > 0. - Convert this column to an integer column so that you can use it in your model and name it
label(this is the default name for the response variable in Spark's machine learning routines). - Filter out missing values (this has been done for you).
# Create is_late
model_data = model_data.withColumn("is_late", model_data.arr_delay > 0)
# Convert to an integer
model_data = model_data.withColumn("label", model_data.is_late.cast("integer"))
# Remove missing values
model_data = model_data.filter("arr_delay is not NULL and dep_delay is not NULL and air_time is not NULL and plane_year is not NULL")
print(model_data.show())
#Results
+-------+----+-----+---+--------+---------+--------+---------+-------+------+------+----+--------+--------+----+------+----------+--------------------+--------------+-----------+-------+-----+-----+---------+---------+-------+-----+
|tailnum|year|month|day|dep_time|dep_delay|arr_time|arr_delay|carrier|flight|origin|dest|air_time|distance|hour|minute|plane_year| type| manufacturer| model|engines|seats|speed| engine|plane_age|is_late|label|
+-------+----+-----+---+--------+---------+--------+---------+-------+------+------+----+--------+--------+----+------+----------+--------------------+--------------+-----------+-------+-----+-----+---------+---------+-------+-----+
| N846VA|2014| 12| 8| 658| -7| 935| -5| VX| 1780| SEA| LAX| 132| 954| 6| 58| 2011|Fixed wing multi ...| AIRBUS| A320-214| 2| 182| NA|Turbo-fan| 3.0| false| 0|
| N559AS|2014| 1| 22| 1040| 5| 1505| 5| AS| 851| SEA| HNL| 360| 2677| 10| 40| 2006|Fixed wing multi ...| BOEING| 737-890| 2| 149| NA|Turbo-fan| 8.0| true| 1|
| N847VA|2014| 3| 9| 1443| -2| 1652| 2| VX| 755| SEA| SFO| 111| 679| 14| 43| 2011|Fixed wing multi ...| AIRBUS| A320-214| 2| 182| NA|Turbo-fan| 3.0| true| 1|
| N360SW|2014| 4| 9| 1705| 45| 1839| 34| WN| 344| PDX| SJC| 83| 569| 17| 5| 1992|Fixed wing multi ...| BOEING| 737-3H4| 2| 149| NA|Turbo-fan| 22.0| true| 1|
| N612AS|2014| 3| 9| 754| -1| 1015| 1| AS| 522| SEA| BUR| 127| 937| 7| 54| 1999|Fixed wing multi ...| BOEING| 737-790| 2| 151| NA|Turbo-jet| 15.0| true| 1|
| N646SW|2014| 1| 15| 1037| 7| 1352| 2| WN| 48| PDX| DEN| 121| 991| 10| 37| 1997|Fixed wing multi ...| BOEING| 737-3H4| 2| 149| NA|Turbo-fan| 17.0| true| 1|
| N422WN|2014| 7| 2| 847| 42| 1041| 51| WN| 1520| PDX| OAK| 90| 543| 8| 47| 2002|Fixed wing multi ...| BOEING| 737-7H4| 2| 140| NA|Turbo-fan| 12.0| true| 1|
| N361VA|2014| 5| 12| 1655| -5| 1842| -18| VX| 755| SEA| SFO| 98| 679| 16| 55| 2013|Fixed wing multi ...| AIRBUS| A320-214| 2| 182| NA|Turbo-fan| 1.0| false| 0|
| N309AS|2014| 4| 19| 1236| -4| 1508| -7| AS| 490| SEA| SAN| 135| 1050| 12| 36| 2001|Fixed wing multi ...| BOEING| 737-990| 2| 149| NA|Turbo-jet| 13.0| false| 0|
| N564AS|2014| 11| 19| 1812| -3| 2352| -4| AS| 26| SEA| ORD| 198| 1721| 18| 12| 2006|Fixed wing multi ...| BOEING| 737-890| 2| 149| NA|Turbo-fan| 8.0| false| 0|
| N323AS|2014| 11| 8| 1653| -2| 1924| -1| AS| 448| SEA| LAX| 130| 954| 16| 53| 2004|Fixed wing multi ...| BOEING| 737-990| 2| 149| NA|Turbo-jet| 10.0| false| 0|
| N305AS|2014| 8| 3| 1120| 0| 1415| 2| AS| 656| SEA| PHX| 154| 1107| 11| 20| 2001|Fixed wing multi ...| BOEING| 737-990| 2| 149| NA|Turbo-jet| 13.0| true| 1|
| N433AS|2014| 10| 30| 811| 21| 1038| 29| AS| 608| SEA| LAS| 127| 867| 8| 11| 2013|Fixed wing multi ...| BOEING| 737-990ER| 2| 222| NA|Turbo-fan| 1.0| true| 1|
| N765AS|2014| 11| 12| 2346| -4| 217| -28| AS| 121| SEA| ANC| 183| 1448| 23| 46| 1992|Fixed wing multi ...| BOEING| 737-4Q8| 2| 149| NA|Turbo-fan| 22.0| false| 0|
| N713AS|2014| 10| 31| 1314| 89| 1544| 111| AS| 306| SEA| SFO| 129| 679| 13| 14| 1999|Fixed wing multi ...| BOEING| 737-490| 2| 149| NA|Turbo-jet| 15.0| true| 1|
| N27205|2014| 1| 29| 2009| 3| 2159| 9| UA| 1458| PDX| SFO| 90| 550| 20| 9| 2000|Fixed wing multi ...| BOEING| 737-824| 2| 149| NA|Turbo-fan| 14.0| true| 1|
| N626AS|2014| 12| 17| 2015| 50| 2150| 41| AS| 368| SEA| SMF| 76| 605| 20| 15| 2001|Fixed wing multi ...| BOEING| 737-790| 2| 151| NA|Turbo-jet| 13.0| true| 1|
| N8634A|2014| 8| 11| 1017| -3| 1613| -7| WN| 827| SEA| MDW| 216| 1733| 10| 17| 2014|Fixed wing multi ...| BOEING| 737-8H4| 2| 140| NA|Turbo-fan| 0.0| false| 0|
| N597AS|2014| 1| 13| 2156| -9| 607| -15| AS| 24| SEA| BOS| 290| 2496| 21| 56| 2008|Fixed wing multi ...| BOEING| 737-890| 2| 149| NA|Turbo-fan| 6.0| false| 0|
| N215AG|2014| 6| 5| 1733| -12| 1945| -10| OO| 3488| PDX| BUR| 111| 817| 17| 33| 2001|Fixed wing multi ...|BOMBARDIER INC|CL-600-2C10| 2| 80| NA|Turbo-fan| 13.0| false| 0|
+-------+----+-----+---+--------+---------+--------+---------+-------+------+------+----+--------+--------+----+------+----------+--------------------+--------------+-----------+-------+-----+-----+---------+---------+-------+-----+
only showing top 20 rowsAs you know, Spark requires numeric data for modeling. So far this hasn't been an issue; even boolean columns can easily be converted to integers without any trouble. But you'll also be using the airline and the plane's destination as features in your model. These are coded as strings and there isn't any obvious way to convert them to a numeric data type.
Fortunately, PySpark has functions for handling this built into the pyspark.ml.features submodule. You can create what are called 'one-hot vectors' to represent the carrier and the destination of each flight. A one-hot vector is a way of representing a categorical feature where every observation has a vector in which all elements are zero except for at most one element, which has a value of one (1).
Each element in the vector corresponds to a level of the feature, so it's possible to tell what the right level is by seeing which element of the vector is equal to one (1).
The first step to encoding your categorical feature is to create a StringIndexer. Members of this class are Estimators that take a DataFrame with a column of strings and map each unique string to a number. Then, the Estimator returns a Transformer that takes a DataFrame, attaches the mapping to it as metadata, and returns a new DataFrame with a numeric column corresponding to the string column.
The second step is to encode this numeric column as a one-hot vector using a OneHotEncoder. This works exactly the same way as the StringIndexer by creating an Estimator and then a Transformer. The end result is a column that encodes your categorical feature as a vector that's suitable for machine learning routines!
This may seem complicated, but don't worry! All you have to remember is that you need to create a StringIndexer and a OneHotEncoder, and the Pipeline will take care of the rest.
In this exercise you'll create a StringIndexer and a OneHotEncoder to code the carrier column. To do this, you'll call the class constructors with the arguments inputCol and outputCol.
The inputCol is the name of the column you want to index or encode, and the outputCol is the name of the new column that the Transformer should create.
- Create a
StringIndexercalledcarr_indexerby callingStringIndexer()withinputCol="carrier"andoutputCol="carrier_index". - Create a
OneHotEncodercalledcarr_encoderby callingOneHotEncoder()withinputCol="carrier_index"andoutputCol="carrier_fact".
# Create a StringIndexer
carr_indexer = StringIndexer(inputCol="carrier", outputCol="carrier_index")
# Create a OneHotEncoder
carr_encoder = OneHotEncoder(inputCol="carrier_index", outputCol="carrier_fact")Now you'll encode the dest column just like you did in the previous exercise.
- Create a
StringIndexercalleddest_indexerby callingStringIndexer()withinputCol="dest"andoutputCol="dest_index". - Create a
OneHotEncodercalleddest_encoderby callingOneHotEncoder()withinputCol="dest_index"andoutputCol="dest_fact".
# Create a StringIndexer
dest_indexer = StringIndexer(inputCol="dest", outputCol="dest_index")
# Create a OneHotEncoder
dest_encoder = OneHotEncoder(inputCol="dest_index", outputCol="dest_fact")The last step in the Pipeline is to combine all of the columns containing our features into a single column. This has to be done before modeling can take place because every Spark modeling routine expects the data to be in this form. You can do this by storing each of the values from a column as an entry in a vector. Then, from the model's point of view, every observation is a vector that contains all of the information about it and a label that tells the modeler what value that observation corresponds to.
Because of this, the pyspark.ml.feature submodule contains a class called VectorAssembler. This Transformer takes all of the columns you specify and combines them into a new vector column.
-
Create a
VectorAssemblerby callingVectorAssembler()with theinputColsnames as a list and theoutputColname "features".- The list of columns should be ["month", "air_time", "carrier_fact", "dest_fact", "plane_age"].
# Make a VectorAssembler
vec_assembler = VectorAssembler(inputCols=["month", "air_time", "carrier_fact", "dest_fact", "plane_age"], outputCol="features")You're finally ready to create a Pipeline!
Pipeline is a class in the pyspark.ml module that combines all the Estimators and Transformers that you've already created. This lets you reuse the same modeling process over and over again by wrapping it up in one simple object. Neat, right?
- Import
Pipelinefrompyspark.ml. - Call the
Pipeline()constructor with the keyword argument stages to create aPipelinecalledflights_pipe.- stages should be a list holding all the stages you want your data to go through in the pipeline. Here this is just: [dest_indexer, dest_encoder, carr_indexer, carr_encoder, vec_assembler]
# Import Pipeline
from pyspark.ml import Pipeline
# Make the pipeline
flights_pipe = Pipeline(stages=[dest_indexer, dest_encoder, carr_indexer, carr_encoder, vec_assembler])After you've cleaned your data and gotten it ready for modeling, one of the most important steps is to split the data into a test set and a train set. After that, don't touch your test data until you think you have a good model! As you're building models and forming hypotheses, you can test them on your training data to get an idea of their performance.
Once you've got your favorite model, you can see how well it predicts the new data in your test set. This never-before-seen data will give you a much more realistic idea of your model's performance in the real world when you're trying to predict or classify new data.
In Spark it's important to make sure you split the data after all the transformations. This is because operations like StringIndexer don't always produce the same index even when given the same list of strings.
Hooray, now you're finally ready to pass your data through the Pipeline you created!
- Create the DataFrame
piped_databy calling thePipelinemethods.fit()and.transform()in a chain. Both of these methods takemodel_dataas their only argument.
# Fit and transform the data
piped_data = flights_pipe.fit(model_data).transform(model_data)Now that you've done all your manipulations, the last step before modeling is to split the data!
- Use the DataFrame method
.randomSplit()to splitpiped_datainto two pieces,trainingwith 60% of the data, andtestwith 40% of the data by passing the list[.6, .4]to the.randomSplit()method.
# Split the data into training and test sets
training, test = piped_data.randomSplit([.6, .4])In this last chapter, you'll apply what you've learned to create a model that predicts which flights will be delayed.
The model you'll be fitting in this chapter is called a logistic regression. This model is very similar to a linear regression, but instead of predicting a numeric variable, it predicts the probability (between 0 and 1) of an event.
To use this as a classification algorithm, all you have to do is assign a cutoff point to these probabilities. If the predicted probability is above the cutoff point, you classify that observation as a 'yes' (in this case, the flight being late), if it's below, you classify it as a 'no'!
You'll tune this model by testing different values for several hyperparameters. A hyperparameter is just a value in the model that's not estimated from the data, but rather is supplied by the user to maximize performance. For this course it's not necessary to understand the mathematics behind all of these values - what's important is that you'll try out a few different choices and pick the best one.
The Estimator you'll be using is a LogisticRegression from the pyspark.ml.classification submodule.
- Import the
LogisticRegressionclass frompyspark.ml.classification. - Create a
LogisticRegressioncalled lr by callingLogisticRegression()with no arguments.
# Import LogisticRegression
from pyspark.ml.classification import LogisticRegression
# Create a LogisticRegression Estimator
lr = LogisticRegression()In the next few exercises you'll be tuning your logistic regression model using a procedure called k-fold cross validation. This is a method of estimating the model's performance on unseen data (like your test DataFrame).
It works by splitting the training data into a few different partitions. The exact number is up to you, but in this course you'll be using PySpark's default value of three. Once the data is split up, one of the partitions is set aside, and the model is fit to the others. Then the error is measured against the held out partition. This is repeated for each of the partitions, so that every block of data is held out and used as a test set exactly once. Then the error on each of the partitions is averaged. This is called the cross validation error of the model, and is a good estimate of the actual error on the held out data.
You'll be using cross validation to choose the hyperparameters by creating a grid of the possible pairs of values for the two hyperparameters, elasticNetParam and regParam, and using the cross validation error to compare all the different models so you can choose the best one!
The first thing you need when doing cross validation for model selection is a way to compare different models. Luckily, the pyspark.ml.evaluation submodule has classes for evaluating different kinds of models. Your model is a binary classification model, so you'll be using the BinaryClassificationEvaluator from the pyspark.ml.evaluation module.
This evaluator calculates the area under the ROC. This is a metric that combines the two kinds of errors a binary classifier can make (false positives and false negatives) into a simple number. You'll learn more about this towards the end of the chapter!
- Import the submodule
pyspark.ml.evaluationas evals. - Create evaluator by calling
evals.BinaryClassificationEvaluator()with the argumentmetricName="areaUnderROC".
# Import the evaluation submodule
import pyspark.ml.evaluation as evals
# Create a BinaryClassificationEvaluator
evaluator = evals.BinaryClassificationEvaluator(metricName="areaUnderROC")Next, you need to create a grid of values to search over when looking for the optimal hyperparameters. The submodule pyspark.ml.tuning includes a class called ParamGridBuilder that does just that (maybe you're starting to notice a pattern here; PySpark has a submodule for just about everything!).
You'll need to use the .addGrid() and .build() methods to create a grid that you can use for cross validation. The .addGrid() method takes a model parameter (an attribute of the model Estimator, lr, that you created a few exercises ago) and a list of values that you want to try. The .build() method takes no arguments, it just returns the grid that you'll use later.
- Import the submodule
pyspark.ml.tuningunder the aliastune. - Call the class constructor
ParamGridBuilder()with no arguments. Save this asgrid. - Call the
.addGrid()method ongridwithlr.regParamas the first argument andnp.arange(0, .1, .01)as the second argument. This second call is a function from thenumpymodule (importedas np) that creates a list of numbers from 0 to .1, incrementing by .01. Overwritegridwith the result. - Update
gridagain by calling the.addGrid()method a second time create a grid forlr.elasticNetParamthat includes only the values[0, 1]. - Call the
.build()method on grid and overwrite it with the output.
# Import the tuning submodule
import pyspark.ml.tuning as tune
# Create the parameter grid
grid = tune.ParamGridBuilder()
# Add the hyperparameter
grid = grid.addGrid(lr.regParam, np.arange(0, .1, .01))
grid = grid.addGrid(lr.elasticNetParam, [0, 1])
# Build the grid
grid = grid.build()The submodule pyspark.ml.tuning also has a class called CrossValidator for performing cross validation. This Estimator takes the modeler you want to fit, the grid of hyperparameters you created, and the evaluator you want to use to compare your models.
The submodule pyspark.ml.tune has already been imported as tune. You'll create the CrossValidator by passing it the logistic regression Estimator lr, the parameter grid, and the evaluator you created in the previous exercises.
- Create a CrossValidator by calling tune.CrossValidator() with the arguments:
- estimator=lr
- estimatorParamMaps=grid
- evaluator=evaluator
- Name this object cv.
# Create the CrossValidator
cv = tune.CrossValidator(estimator=lr,
estimatorParamMaps=grid,
evaluator=evaluator
)You're finally ready to fit the models and select the best one!
Unfortunately, cross validation is a very computationally intensive procedure. Fitting all the models would take too long on DataCamp.
To do this locally you would use the code:
# Fit cross validation models
models = cv.fit(training)
# Extract the best model
best_lr = models.bestModelRemember, the training data is called training and you're using lr to fit a logistic regression model. Cross validation selected the parameter values regParam=0 and elasticNetParam=0 as being the best. These are the default values, so you don't need to do anything else with lr before fitting the model.
- Create
best_lrby callinglr.fit()on thetrainingdata. - Print
best_lrto verify that it's an object of theLogisticRegressionModelclass.
# Call lr.fit()
best_lr = lr.fit(training)
# Print best_lr
print(best_lr)
#Result
LogisticRegressionModel: uid=LogisticRegression_49e43ef278e1, numClasses=2, numFeatures=83For this course we'll be using a common metric for binary classification algorithms call the AUC, or area under the curve. In this case, the curve is the ROC, or receiver operating curve. The details of what these things actually measure isn't important for this course. All you need to know is that for our purposes, the closer the AUC is to one (1), the better the model is!
Remember the test data that you set aside waaaaaay back in chapter 3? It's finally time to test your model on it! You can use the same evaluator you made to fit the model.
- Use your model to generate predictions by applying
best_lr.transform()to thetestdata. Save this astest_results. - Call
evaluator.evaluate()ontest_resultsto compute the AUC. Print the output.
# Use the model to predict the test set
test_results = best_lr.transform(test)
# Evaluate the predictions
print(evaluator.evaluate(test_results))
#Result
0.7123313100891033