Chris, I not completely sure I'm following you, (your screenshot is confusing), but I'll take a shot in the dark. Tableau has a default of 6 for number of columns. If you go over 6 T combines fields. Go to Analysis\Table Layout\Advanced and you'll get this:
Change them to 16 (the maximum), and see if this helps.
Shawn. Thanks for the reply! (that was a long wait, what have you been up to? haha)
Yes. It does look confusing, because I had to copy several screen shots into the image to convey all the information. Unfortunately the issue is not a limitation with the GUI. Its more serious.
Just notice the teal colored box at the top shows these noted labels. This is the objective--alternating two different data sets:
t6, t12, t6, t12, t6, t12, t6, t12
These data sets represent two blocks of time series data from time t0-t6 and t0-t12. However, two of the columns appear to have combined values (t6+t12). These are indicated with the two arrows from the teal box towards the "pill's hover data that I pasted into the screen shot. Those are outlined in red and labeled with #1. In short, that's the problem. Why do these two pill pop-ups show the measures the way they do? The rest of my reply is just more explanation of the screen shot.
I'll skip describing #2 right now, and direct your attention to #3. If you'll recall the data set note above, t6 data is contained in t12. Therefore t12 should always be a larger value in the graph. However, #3 shows these bar graph sizes reversed, or something else, or what I'm thinking is happening: t6 = t6+t12 in those two instances!?!
The trouble is this 'combined' effect is not a mistake in drag and dropping pills over from the side pane (#2). Tableau seems to automatically combine these data. For reasons I don't understand, Tableau seems to be combining these two measures automatically only in these two instances. Are they somehow linked in the background?
Chris, given all that, and that you don't seem to be able to share the workbook, I suggest you open a support ticket. They are good at keeping your data confidential. Someone is going to need to see the workbook to understand the behavior you are describing.
What do you mean? Share the workbook? Where do I do that?
Thanks for your help.
I'm confused how this topic about sharing my workbook started--whatever. I tried to get approval to share extracted data set, but I've not heard back from the content owners. Looks like sharing the workbook is a no go. As I said in my last email, I'm not sure how to share my _workbook_. WHich is to say not that I don't know how to post an attachement, as I clearly added to my OP, but how to share a workbook in a useful way. As near as I can help, it requires zip archive of the tableau file and the csv data (which as I explained, I don't have permission to share).
Unless there is some other way to share tableau file with just the data necessary to make the chart shareable for inspection, I can only offer the screen shot. The simplest explanation of that image starts with hover of two pills showing a string which suggests two measures are combined (or joined?), when only one measure is expected. This happens only on two of the eight column pills.
Sorry I misunderstood your level of expertise in using the forums. Therefore I won't mention that mocking something up using the Superstore data set and posting a package workbook is your best chance of getting a meaningful answer to your question -- as you're already aware of that.
OK. I took a look at the Sample data in Superstore. Unfortunately this doesn't reflect the situation very well. You see I'm connecting to four data sources.
Data source one and two, lets call it NY Superstore, I have two files:
1) sales on 12-01-2014 at noon.
2) sales on 12-01-2014 at close.
My third and fourth data sources, lets call CA Superstore, where I have another two files,
3) sales on 12-01-2014 at noon
4) sales on 12-01-2014 at close.
For each "store" I perform a left join on User ID (UID) where the "at close" file is on the left (the file with the complete day's sales) and the file "at noon" file is on the right (all UID will be included, because "at noon" is a sub-set of "at close"). This shows the change in one store from mid-day to end-of-day.
Plus, every file has these two measures:
a) units sold
b) gross receipts
Finally, to complete the comparison between my data and the Superstore example, let's say our store is so amazing, some of the NY shoppers are flying to CA to continue shopping after NY store is closed (they have a really fast jet). This is the relationship between NY and CA in this fictitious rendering of my data set.
Sorry, I don't know how I can recreate the trouble using the Sample data since it involves four files and not one.
I don't know how I could share my workbook using Superstore as the example set.
SO....If you were looking at YOUR workbook, and put the mouse over each of the column pills to see which measure was being shown and saw these (see attachment) what would you think?
tableau_v2.jpg 130.0 KB
Sorry I didn't fully read your last reply. So ignore my last response. I really don't know how to help you. I can only suggest you open a support case.
No worries. I don't have a support contract. This is a student lic. No Joy. Peace