Alteryx vs SQL

[ThizViz]

I'm trying to build a case for adding Alteryx to our toolset. We're a SQL shop and for those of you who were using SQL and now have Alteryx, can you give me some examples of advantages? Not so much for things you can't do in SQL anyway, or not well, like spatial or predictive. But how has Alteryx improved processes previously done with SQL?

Thanks!

[ThizViz]

If I'm not mistaken, it could be as simple as TONUMBER([Field]). But I admit I'm out of practice :-)

[papalow]

@Treyson 

Great visual.  

Thanks for sharing it.

[garthn555]

In Oracle you can use either the translate or regexp_replace functions. You can also easily implement your own code as functions, as if they were part of the database native set. Is there a way to do this in Alteryx, so that they can be used eg in the formula tool? Still early days for me in learning Alteryx...

[Treyson]

@garthn555, 

Can we clarify this ask real quick? Are you saying that there is a native oracle function called Regexp_Replace? If that is what you are asking there are two ways to go through it. There is simply REGEX_Replace, which exists in the formula tool. You can also use the Regex tool and select the replace function.

Or are you asking about using the IN-DB tools to apply these formulas? That's a different ball of wax

Party on Garth.

[garthn555]

Sorry, maybe I wasn't very clear. I just meant that an Oracle solution (not using Alteryx) could use regexp_replace, or other methods. What was originally posted is not a good method and therefore not a good comparison to Alteryx.

My post was really aimed at finding out whether the standard function set in Alteryx can be expanded (via Python coding perhaps?).

I haven't got into the in-db stuff yet - I will start on it soon

[garthn555]

The real advantage is that non-coders can develop solutions using it.  There is nothing that Alteryx does with the basic tools that SQL cannot do. But most business and data analysts aren't capable of writing advanced (eg analytical) SQL.  Alteryx (by design) breaks everything down into simple and easily understood steps. SQL plus a host language such as PL/SQL or T-SQL is both more complex and powerful, but there isn't much that can't be done by several simple (ie Alteryx tools) steps.

Macros?  no, not an advantage. Eg Oracle packages are a standard method of development

Visual? yes - Alteryx definitely makes it easier to visualize what a solution does

Ease of development? for someone relatively unskilled in advanced database development (ie code generation), Alteryx is much easier

Performance? database/SQL by far, there is no way Alteryx will handle the volumes I have worked with on engineered platforms

Interfacing? not much of a difference, except that it seems easier to expand the database capabilities (eg for BSON)

[JohnJPS]

If you're extremely handy with PL/SQL, e.g. you "think in PL/SQL," then perhaps Alteryx will seem like it offers you rather little.  I've been a developer for decades and early on with Alteryx, I found it easier to do things in either SQL or even raw C# against flat datasets (or raw R or Python), but over time, especially for quick and dirty analysis, I have found it quicker and easier to drag a dataset into Alteryx and start slicing/dicing using Alteryx tools.  Pivoting in raw code has always been annoying to me; cross-tabl/transpose in Alteryx is almost as easy as thought.

With loops, I do still often lean on R, since I selfishly find coded loops more elegant than an iterative macro. That's just the programmer in me, though, and happily, R is built into Alteryx, so I can insert when and if needed.

For anyone not already familiar with structured programming or SQL scripting, Alteryx is a very powerful self-service tool.