Thanks for the reply.
I've checked those web links earlier and tried them but could not figure out how to get the coordinates.
And to my requirement I need to build a network where one Doctor is affiliated to other Doctors and the other Doctors to their affiliated Doctors. It should be a One-Many(1st instance in the network), Many-Many(2nd instance in the network) network diagram connecting all the doctors.
How do we create data and also their respective coordinates and build the network graph.
Any possible solution will be more than appreciated.
1 of 1 people found this helpful
You’ll need to use another tool to create the network. A couple are Gephi http://gephi.github.io/ and NodeXL http://nodexl.codeplex.com. Both of these tools have the ability to create network graph visualizations, so you might not need Tableau at all, the reason why I picked them is that a) they are popular and b) have the ability to export the graph data so you could open that in Tableau.
How do I find the Gephi export data function? I can't find it...
I find it remarkable that Tableau offers no help in this area. I went down this path a year ago in an effort to represent node/edge for call detail records. I gave up, Tableau simply doesn't support this kind of functionality, I ended up using SAS Visual Analytics but that was a very expensive option. Sorry I couldn't be of any help, but my advice to you is to look beyond Tableau for this kind of functionality.
1 of 1 people found this helpful
I have decided to create my own solution, which can also be used for real production environments. It involves using Python for the data preprocessing:
I look forward to checking this out. Thank you for your contribution.
Raja, to answer one of your questions, you don't "get" the coordinates from anywhere - you make them up by trial and error.
Unless you're an R or Python programmer, there are very limited options for analyzing a given data set to extract the relationships between column values, and then to create an output that Tableau can use to draw your nodes and edges.
Yes, as Michael stated above, the co-ordinates are determined manually and you can choose any set of points and place the nodes in dispersed manner in Tableau view. Tried to get the information from the excel sheet provided here -> Build Network Graphs in Tableau - Clearly and Simply and blended (sliced and diced) my data to fit in the model stated here, and was able to plot the network graph.
I am wondering if I can get to know how we can create LineX, LineY and CircleY then it will solve my purpose and will be thankful.
LineX LineY CircleY
Nikhil, as I noted above in this thread, you make them up. Their values are completely up to you.
If you imagine a grid with an x-axis of 1,000 "ticks" and a y-axis of 1,000 "ticks", you simply assign x & y values to each point you wish to display within that 1,000 x 1,000 grid. It's not more complicated than that.
You'll have to play around with the placement of your points but just start by plotting the points to see what it looks like, and then changing the x or y coordinates for a point as needed.
Hi, I have prepared a easy solution but with the help of Alteryx and R. But it can be done with any open source or SQL language. Follow the link, Alteryx Love Tableau: Data Preparation for Network Graph - Data Police Inc.
I followed the post and was able to create my network graph on tableau. You can see my python notebook of the nework graph with tableau in this GitHub.
I believe you can do more with pos=nx.shell_layout with centers but I haven't tried.
Hope it helps
Thank you for getting this very clear coordinate generator together in Python! I had been searching all over and this looks like just what I need to get my data ready for displaying in Tableau!
I combined the X,Y generation logic from Use Python to create network graphs in Tableau – Daan Tor
and then the Clear and SImply directions Build Network Graphs in Tableau - Clearly and Simply
and was able to create the following four visualizations to show the same integrations data set. I simply needed to know the Source, Target, and a Weight (I used Number of Integrations between systems). In the data example there are 4 Source Systems and 84 total systems. The Source Systems are Targets in some cases.
All Source Systems
Filtered to 1 Source System
Filtered to 2 Source Systems
Filtered to 3 Source Systems
The narrative can get very interesting when you start looking at animation - but I have not gotten there yet (I am technically a Tableau Newbie!)