Apache Spark™ Tutorial: Getting Started with Apache Spark on Databricks

Overview

The Apache Spark DataFrame API provides a rich set of functions (select columns, filter, join, aggregate, and so on) that allow you to solve common data analysis problems efficiently. DataFrames also allow you to intermix operations seamlessly with custom Python, R, Scala, and SQL code. In this tutorial module, you will learn how to:

Load sample data
View a DataFrame
Run SQL queries
Visualize the DataFrame

We also provide a sample notebook that you can import to access and run all of the code examples included in the module.

Load sample data

The easiest way to start working with DataFrames is to use an example Databricks dataset available in the /databricks-datasetsfolder accessible within the Databricks workspace. To access the file that compares city population versus median sale prices of homes, load the file /databricks-datasets/samples/population-vs-price/data_geo.csv.

%python
# Use the Spark CSV datasource with options specifying:
# - First line of file is a header
# - Automatically infer the schema of the data
data = spark.read.format("csv") 
  .option("header", "true") 
  .option("inferSchema", "true") 
  .load("/databricks-datasets/samples/population-vs-price/data_geo.csv")

data.cache() # Cache data for faster reuse
data = data.dropna() # drop rows with missing values

View the DataFrame

Now that you have created the data DataFrame, you can quickly access the data using standard Spark commands such as take(). For example, you can use the command data.take(10) to view the first ten rows of the data DataFrame. Because this is a SQL notebook, the next few commands use the %python magic command.

%python
data.take(10)

Output of the DataFrame take() command

To view this data in a tabular format, you can use the Databricks display() command instead of exporting the data to a third-party tool.

%python
display(data)

Displaying a DataFrame in Databricks

Run SQL queries

Before you can issue SQL queries, you must save your data DataFrame as a temporary table:

%python
# Register table so it is accessible via SQL Context
data.createOrReplaceTempView("data_geo")

Then, in a new cell, specify a SQL query to list the 2015 median sales price by state:

select `State Code`, `2015 median sales price` from data_geo

Or, query for population estimate in the state of Washington:

select City, `2014 Population estimate` from data_geo where `State Code` = 'WA';

Visualize the DataFrame

An additional benefit of using the Databricks display() command is that you can quickly view this data with a number of embedded visualizations. Click the down arrow next to the