Can I change the data type of an extract so Tableau will still utilize it, yet it will be untouched by "REFRESH ALL EXTRACTS" ?
I'm using Tableau Desktop 9.2.7 with no access to Tableau Server. My issue is as follows:
- I've created a Tableau Extract from an ODBC source. Let's call it COMPANY RESULTS (OLD)
- I'm expecting a bunch of new data, Let's call that COMPANY RESULTS (NEW)
- I want to compare COMPANY RESULTS (OLD) to COMPANY RESULTS (NEW) and determine where significant changes have occurred
- Here's where it gets tricky (at least for me): Once the new data comes in, it effectively replaces the old data, wiping it out as if it never existed. So I need to preserve the old data, then import the new data under a different name. I'm aware that I could copy the extract and manually refresh the one I want, my fear is I'm going to forget this little step and accidentally hit "REFRESH ALL EXTRACTS"-- causing both extracts to reflect the same values and rendering any comparison useless.
Can I take my existing extract of old data and turn it into a data type that Tableau will play nicely with, but will remain unchanged if I accidentally hit "REFRESH ALL"?
Because of the size of the extract, I doubt I would be able to manually copy/paste underlying data into a .csv file: a systemic method is needed.
I'll be happy to attach a .twbx if you think that will help, however I'm hoping this is more of a procedural question. Thank you for your help!
--Michael
The short answer? No. You can't do this. You're trying to turn "Refresh" into "Archive", which is not it's function. It's function is Refresh
The longer answer is that you CAN do this, but it's unnatural because you're making the tool do stuff it wasn't really designed to do. Here are a few things you might try:
- Create "old" extract". Make a COPY of this TDE ("old extract copy") , put it aside.
- Refresh your extract - "old" extract is now "new" extract and doesn't have the "old extract" historical data in it anymore
- Choose Extract | Append Data From File and grab data from your "old extract copy")
....Now you have 2x the data - new and old.
The other thing you could try is to leverage Data Blending between your "old extract copy" and "new extract" on some sort of key and then comparing rows which share the same key...but this won't help if you are some records which don't MATCH others...
Good luck!
