Alteryx Community - Best Practices

[judyshere]

When separating data into columns and rows, which delimiters do you most commonly use?  Is it best to stick with commas?  Can you run into trouble if you use just one type of delimiter?  If you use many?  What is the "best practice"?  Thanks!

[danilang]

Hi @judyshere 

Using common characters like commas can lead to problems when delimiting data that contains the character.  For example, "John Smith, Esq." is a name that contains a comma and so the entire string would need to be enclosed in quotes.  Many regional settings use commas as either the thousands delimiter or as a period so these can be problematic as well.   The Data Output Tool can deal with these cases by using smart quotes, but if you're concatenating data that will be used further in the same workflow, a more reliable character to use is the pipe character "|".  It's almost never found in text, unless you're parsing programing manuals. 

If you're outputting data to be used by an external program, then you have to follow the rules of the external program as well.  Most programs that read delimited files have a way to specify what the delimiter is if it's not a comma, but some older ones require commas.  

The only case that I can think of to use more than one delimiter is if you're nesting data, i.e. 1|Many#Brown|555-555-5555.  Here the second delimiter "#" is used to split first and last name   In this case, use another uncommon character such as the "#", though it's becoming more common. 

Dan