Most of the time an Alteryx workflow starts with an input tool set to ingest some file or stream data from a database. However, sometimes you would want to download a flat file from the Internet, probably from the URL calculated dynamically, to use it as one of the input datasets. The standard Input tool does not accept an URL as an input.

So, suppose, we need to download fresh data about COVID cases in NSW every time we run the workflow. The URL of the dataset is

https://data.nsw.gov.au/data/dataset/aefcde60-3b0c-4bc0-9af1-6fe652944ec2/resource/5d63b527-e2b8-4c42-ad6f-677f14433520/download/confirmed_cases_table1_location_agg.csv

Step 1. Enter the URL through text input (note: you can dynamically calculate/update it, if necessary, upstream).

Step2. Run the Download tool with this URL/URLs.

Now all your data is in one cell. Unlike the Input tool, the Download will not transform CSV into rows and columns automatically. Therefore…

Step3. First, use the Text-To-Columns tool with “\n” as a separator and the “Split to Rows” option (I know, the tools is called “split to columns”, but…) 

You will get a table with rows separated (but all columns are still concatenated in one field). 

Step 4. Finally, use another Text-To-Columns tool to divide the columns on the “,” divider (with “Split to Columns” option). And do not forget to increase the number of the columns in the tool

Your job is effectively done here.

You can use the Dynamic rename to pull the column names from the first row, and apply the Select to remove unnecessary columns, but otherwise your data is ready for analysis.

Eugene Kutilov
Author: Eugene Kutilov

Eugene has a background in science with a master’s degree in quantum physics and decades-long experience in technology marketing. Eugene’s technology toolbox includes Alteryx, Tableau, PowerBI, SQL, Python, R, GCS. He is a Tableau Certified Data Analyst, Microsoft Certified PowerBI Data Analyst (Associate), and Certified Alteryx Designer Advanced Specialist. He is also a Certified Data Scientist by Datacamp.