What’s everyone’s take on Alteryx?

Hate it as I feel like it creates bad habits. It is a crutch for people who dont want to bother to learn SQL. Then they refuse to learn SQL it seems. The redeeming feature is the prepackaged geospatial data sets.

Without the Geospatial data sets I'd rather have a real programmimg language (R or Julia) plus a database (Hive, spark, SQL server, postgis etc depending on data volume).

Love it

Incredible for visual programming. Personally it was a stepping stone to learn data science

Love it. It’s amazing for prototyping.

Absolutely love it. It’s more than just ETL. It dies spatial, it does modeling, it connects to datasets previously only dreamt of and if you know how to work it can automate the life out of your mundane tasks.

I don’t like to tout software but this is an exception

Alteryx can't even do a for loop... Seriously how can you love this software?

If you arent working with geospatial, have small to medium data sets, and must have a workflow based tool id recommend KNIME.

It is free and has sensible features like loops, try catch error trapping, reasonable integration with R, Java and Python, and you don't rerun every data set from the beginning of the workflow. It also works closer to a REPL than alteryx.

Side note but I expect you could build spatial capabilities into KNIME with minimal effort.

Thank you all for the feedback. It looks like I have some options to explore.

I have no need for geospatial analysis as I am using arcGIS. I am more looking for something to handle analyses of larger data sets that are beyond the capabilities of excel. With VBA I’ve created analysis templates that can run a pretty complex multipart analysis on 7 million+ records easily. My difficulty lies in exploring the data due to required changes in code. I’m not versed in python though I would like to learn as the arcGIS software has an interface for it.

D1 you can cache datasets in Alteryx so you don’t need to rerun them. Another way is to write them to Alteryx yxdb files which are super small due to its incredible compression and use that as you datasource while testing things out.

It sure what you mean about looping but you can definitely create macros to iterate runs which will loop through a set of criteria. Not to talk about creating an analytical app that the user just uses a simple form interface enabling them control inputs and outputs.

And it does have error trapping. You just need to enable it and configure what kind of errors you want to capture. Can even send you an email of the errors too
I will check out knime though. Haven’t tried it.

R has a number of spatial packages for querying spatial data. SQL Server and PostGIS are good options as well.