Hi Anand Bale...
I'm not sure what this is... I'm not sure why teenagers insist on "cactus diving", either, but my advice is the same: just because you can do it doesn't mean you should do it. Please don't do this (especially if you consider yourself a Tableau Developer).
It's pretty and it's fancy, yes.
It's nearly impossible to read:
Primarily because there are too many similarly-toned colors squished in side-by-side.
Can you REALLY tell the difference between these two pairs?
Remember that color-blind users won't be able to see many of these colors.
You don't have a color legend, and your colors don't seem to be consistent with volume. Again, this makes it difficult to figure out.
You have "orphan units," adding to the confusion. Is it reasonable to expect a user to connect these two?
The label ("545") is off-center (easily fixed)
I know it looks like I'm just trying to give you my opinion of what's wrong with the viz, rather than how to make it. But instead of this Franken-viz, why not try a Horizontal Gantt-bar chart of non-aggregated values?
It might look like this:
Measure A ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Measure B ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Measure C |||||||||||||||||||||||||||||||||||||||
Measure D |||||||||||||||||||||||||
Measure E ||||||||||||||||||||||
Measure F ||||||||||||| Total Widgets: 545
Make sure you have a clear axis, or better yet-- add Reference lines for each measure showing the totals so the user doesn't need to count the incidents.
Good luck to you! --Michael
Many Community members have strong opinions about Donut-y graphs.
I agree with everything Michael Hesser said. If this is an internal project then talk to them about what Michael said. If this is a consulting engagement then still talk to them, but if it *must* be done this way you can do it - it's hard, it will take time, but it can be done.
Perhaps it's a stadium-seating chart?
My supervisor initially thought it might be a seating chart for a symphony, but it's unlikely a symphony would have that many instruments.
It would be good to know if this is representing a physical object (a map), vs. a depiction of measurement (a graph). That could change how you approach the problem. Thoughts?
I thought the same - stadium, senate, parliament, something like that would mean it MUST be done that way. Orchestras can be that big, my kids were in something called Schools Spectacular recently (Homepage - Schools Spectacular ) 2700 in the choir, 2300 dancers and 80 in the orchestra - HUGE !!!
That's why I say ... from a "best practice" point of view then there's a few better choices that should be used. From a "we have no choice, it must be done this way" then yes, collate the data, format it the right way and it's done
1 of 1 people found this helpful
This is pretty rough (it only took 15 mins to do though) :
I took the original image, zoomed it up 300% and then used to http://drawingtool.powertoolsfortableau.com/ to create the points.
I would suggest using maybe a 500% zoomed image, be very precise and you will have an awesome image
I don't know how to name each point properly, but if it makes sense I'd do each arc (you can see by my numbering that I didn't do that, and you can get a bit lost in the process)
Chris McClellan i can probably name a dozen not the whole 545 . . I was never a journo . I recently came across a chart like this . and thought if we can do in tableau . And you have showed the way in no time.
Cool challenge to plot along circles. I did not follow requirements exactly but just wanted to see how easily you can plot something like that:
that was fun
10.5 wb attached
dots on circle.twbx 25.1 KB
That's awesome work I think for the required viz they need to ID each point so they can assign a colour and other information (sitting member maybe ?)
But still impressive that 2 rows of data can do that
Just speaking on Anand Bale's behalf and what we've learned in this thread - this is seats in the Indian Parliament, so the first step was to check that this is doable in Tableau, but the next step is to get the real data on to the viz and I'll add the extra complexity of being able to change the seats (when there's a new election) but that's just another dimension in the dataset.
So although the calculations using COS (etc) and hexbinning the data is REALLY COOL and great to see, each point needs to be identifiable so it can be joined with the real data in order to get the real data mapped to the "seat" properly.
I don't want to invent the seat data because I don't understand the Indian parliament at all, but hopefully Anand Bale can help us with more data so a full solution can be achieved.
Either layout could be tweaked
and exported to a CSV for a hardcopy.