Hi all,
We were working through some ideas with @BenG and @KatieH about how we can streamline the process of understanding data at the point of bringing this into Alteryx.
We currently have a visual way to see columns that have leading whitespace or null (with the visualitics that have been added to the browse tools) - we can take this further:
- After any input tool, alteryx looks at the data to see what can be done to help cleanse:
- for columns that look like they have dates - check cleanliness and offer to add the required tools to get these into a clean date format
- for columns that look like ZIP codes - offer to check vs. US ZIP codes and cleanup
- For columns that have leading or trailing whitespace - offer to add a tool to the canvas to clean this up
- For number columns that have nulls - offer to add a default
- For columns that have strings that contain dates - quick wizard for any anomalies and then add in the tools to automate this cleanup next time.
- For columns which are XML - offer to check this for XML errors
- For any CSV files - check for rows that have different number of columns - and intelligently sideline these
Putting this wizard immediately after input tools, which would then automatically add tools to the canvas to automate the cleanup that we'd agreed during the wizard - we can dramatically reduce / eliminate the cost of the standard data prep tasks.
NOTE: This is exactly how Watson Analytics works when you bring in data, and it makes the process much quicker.
