Reading a file into a dataframe using PySpark in Databricks

By

The very first thing we need is a file. For this demo, this is our sample file : 

A comma-separated file with 3 columns and 3 rows.

  • So now that we have that, the first step would be to actually upload the file over to databricks.
  • Get the cluster running.
  • Go to Data tab
Click on "Create Table"



Drag the file to the specified space.


  • At this point, the file will be uploaded to the DBFS storage. After this, we can work with the file.
Copy the location of the file. (Here: /FileStore/tables/SampleData-1.csv)



  • Now, let's get back to the notebook to work with the file. 
  • Creating a dataframe from the file
dataframe_var=spark.read.option("header","true").csv('/FileStore/tables/SampleData-1.csv') 
  • Display the dataframe  

  • We can see that the file data is successfully displayed in the dataframe.


We will take a look at how to load files to tables in the next blog.


Comments

Post a Comment

Popular posts from this blog

Creating tables from dataframes

By
Image
 Continuing from our last post: Reading file into dataframe Let's create a table from the dataframe: Dataframe.write.saveAsTable("Database.tablename") We need to remember that we are creating a managed table with the above syntax, and to create an unmanaged table we can add the path option.   Managed Table Unmanaged Table Now we can Query the tables using SQL. Reference for the difference between managed and unmanaged tables: Click here  
Read more