Multi Level Dendrogram Chart in Tableau| Hierarchy Chart| Tableau Part II

[@renansampaio3640]

This is AMAZING! I'm fascinated by the good explanation in this video. All the steps can be done easily! Thank you very much!

[@anandsharan8286]

I am surprised as to Why there are so few Likes and fewer subscribers on such a brilliant Article. Thank you Gurpreet for the hardwork and patience to explain in both your videos of Dendogram chart (Single and Multi level). Cheers and e-peace

[@DataVizCanvas]

Thanks Anand for your kind words 🙏 Stay tuned for more such videos 😊

[@richadwivedi1984]

I just love it. Going to tryout today. Will connect if I need any help. Thanks for this graph.

[@DataVizCanvas]

You are welcome. Please let me know how did you go. Feel free to reach out if you need any help

[@JoeOh-z5r]

Great video! Is there a way to keep the category order static so that applying filters does not change the order? Alternatively, is there a way to reposition the sub-category chart based on a filter update where categories are added or removed?

[@roes5654]

Thank you!

[@iuliastefan4636]

Hey Gurpreet, this is literally the most helpful data viz guide I've come across. I was able to follow the very clear instructions step by step and create the multi-level chart. Thank you so much for putting together this amazing resource! I did have a couple follow-up questions + tweaks I need to make for my use case and was wondering if I could get your thoughts: 1. I noticed you divide your Sales numbers in the example by 2 -- why is that? I couldn't see a difference in results if I did/did not include this so I ended up removing it to ensure the # displayed in the chart is correct (otherwise it was showing it as half its value given the division by 2) 2. For my 'sub-category' section of the chart on the very right, I can have up to hundreds/thousands of results and am trying to figure out a way to restrict this to make it readable. I tried just applying a Top N filter to the data and limiting it to the first [25] results based on rank, but the chart looks weird when I do this. It does seem to restrict the # of results, but it keeps them in their original place on the chart vs re-centering the chart to show them symmetrically. It's a bit hard to describe but hopefully that makes sense. Do you have any examples of creating these types of charts with many potential results and how to limit them to something legible (eg taxonomies that contain more than a handful of labels)? Thanks again for all your help!

[@iuliastefan4636]

Nevermind on #1 - realized you're dividing by 2 because you're doubling everything with the introduction of the path cross join. My example is a bit different because I am counting unique users vs summing, so I had to do my partitioning a bit differently but overall principles still hold :)

[@DataVizCanvas]

@@iuliastefan4636 Hi, Firs of all i would like to thank you for supporting my channel. Yes , you are right. I am dividing it because of two values used in path field for data densification. In regards to your second question, this chart doesn't support too many drill down values but you can filter them out using data source filter or context filter. However there is another option to create Org chart and can be checked using below link: Tableau Dashboard: public.tableau.com/app/profile/gurpreet.singh2669/viz/FamilyTreemockup/FamilyTree Video Tutorial : kzbin.info/www/bejne/eGGsqXd-baigbpI Hope that helps

[@iuliastefan4636]

Thanks @@DataVizCanvas for the quick response, much appreciated! Also thank you for sharing that other example with the multi-level drilldown, very cool! However I think this is a bit different from what I would need for my use case. Basically, I don't have too many levels - only 3 max - so your original tutorial above works for that. The additional complexity is that for me, the 3rd or lowest level can contain tens/hundreds of potential values vs 15-20 as you show in your example above (in other words - imagine instead of 15 potential subcategories you had 300 you were trying to show). I am trying to figure out how I can either introduce a scrollbar to the results or just limit to showing the top 15 results with an option to click to see more. Otherwise if I just simply add in all the values to the chart as-is, everything gets squished together into the y-axis range I have manually specified..hope that makes sense, and again appreciate your time and input!

[@jameskim6344]

Amazing. Thank you.

[@DataVizCanvas]

You are welcome 🙏🏼

[@MoooMooMoo]

another fantastic video thank you!

[@DataVizCanvas]

Thanks Wendy😊

[@lama7559]

Thanks, it helped me 👍 What if I have many values and I just want to show the top 10 values in each level ??

[@DataVizCanvas]

You could use a filter. Example: In the first level of dendrogram where I'm showing categories only. Use Categories in the filter shelf. Then edit the filter and you will see 4 tabs in the pop up window i.e. General, Wildcard, Condition & Top Select the Top one Then select the radio button for "By Field" Then select top 10 (Or. what ever is your requirement) Then select "sales" measure and aggregation of "Sum" then click ok All done :-) do this in all the levels to achieve desired result

[@lama7559]

@@DataVizCanvas Thanks a lot. I tried to use the top filter and it works well with the first level, but I have a problem with the second level, the top filter doesn't show the actual top 10 values, I don't know why. Is it possible for you to have a look on my workbook?

[@DataVizCanvas]

@@lama7559 This is happening because of order of execution of filters. In this example it filters the Sub category filter for top 10 values is executing first and then your @Category calculation is running which is a boolean value. In order to overcome this please set the @category Calc to context filter. You can right click on the @category cal filter and select Context filter. This will give you the desired result. Hope that helps :-)

[@PriteshRPatel-lr5uh]

Another Fantastic video. I'll be thinking about how to add a 3rd and 4th level.

[@DataVizCanvas]

Thanks Pritesh. You can add the 3rd and 4th level in the same way. Please try and feel free to reach out if you need any assistance.

[@santanaN77]

any video for create a 3-level dendrogram chart?

[@houselan7636]

Thank you so much for this tutorial! It helped me a lot. I created a three level dendrogram and it worked perfect. There is only one problem. It seems exist a system error when do the color coding for the last level. Every time when I set the different color for a specific category, the entire group changed to that new color. I followed all your steps. I don't know what is wrong. Please give me some tips on this. Supper appreciated.

[@DataVizCanvas]

Hi, Seems like your colour selection is overwritten by some parameter when you select a category. Check all the parameter actions and make sure you have colour marks set up for category and subcategory as well in the same theme. Please try that and let me know if the issue still exists. Please share link to your workbook if the issue still exists and i can investigate further. Cheers!

[@Karavaki]

@14:11 how that can be automatical ? I have plenty of categories and the user will filter so having a static axis , is not helping. any chance it can be automated?

[@josephubbarapu8885]

Perfect Loved it. Thanks for this Tutorial. "Hey but I can see the sales values are showing wrong. All the sales values which are showing in the dashboard are half of the original sales values. please check once any formula is taking "'/2'" Example Original sales for Furniture -> Chairs = $ 656,898 but in the dashboard it shows $ 328,449 exactly half of the sales. Please give an update on this."

[@DataVizCanvas]

Hi Joseph, In order to create this dendrogram chart we need to use data densification technique. And for that we have joined the sample super store data base with and excel spreadsheet with 2 values that is 0 and 200 You can check it in the part I of the video at 2:08 min Link to video: kzbin.info/www/bejne/hJqTmmOMn7Sod9k Due to joining it with additional two rows , i have to divide all the values by 2 to get correct sales number. If you connect to sample superstore data set and check the values of sales for chairs, it will be $328,449 Hope that helps. Please feel free to reach out if you need more information. Cheers! Gurpreet

[@vipinpalsingh9760]

thanks for the great video@data Viz......i had a query how you manipulate and get the correct value of "0.062".Is there some calculation based formula. If yes...Pls Share that also.

[@DataVizCanvas]

Thats a great question. However its just by hit and trial to get the proper allignment.

[@Cudicut]

Awesome video! I have two questions. 1) would this be difficult to make into 4 levels/do you have any tips for doing so? And 2) would it be possible to either not show the later levels until a selection is made in the first/prior level?

[@DataVizCanvas]

Yes you can do it to any number of levels but it will get more complicated as you increase the number of levels. However i would strongly recommend to keep it till 3 levels and at max 4 if required. Also you can show the levels based on the prior selection by adding more calculation and parameter action. Please try and feel free to reach out in case you have more questions. Cheers!

[@Lau52085]

Hi! I am starting to use Tableau for the first time. I like this visual. I have a question, can you use this graph to see how many investigations cases were investigated per region and then trickling down per area of service? I would be interested to see the correlation of membership population to total number of investigations per region--->area of service....

[@DataVizCanvas]

Hi Laura, Yes, you could use this visual to achieve same results. First Level can be "Cases by Region" and then once you select individual region, it will drill down to individual areas showing number of case in each area. Please try and feel free to reach out in case of any issues.

[@Lau52085]

@@DataVizCanvas oh my god thank you so much! I will try first and then if I get stuck, maybe you can help me?

[@gururajb.m2472]

This is amazing viz, very informative... Can you please make a video about creating a Circular Dendogram chart?

[@DataVizCanvas]

Thanks 🙏🏽 for your feedback. I will add that to my list of future videos.

[@wasekar]

Nice Video and explanation is there video of Muti level Sankey diagram ?

[@DataVizCanvas]

Hi 👋, Multilevel Stanley is in my to do list and will publish it soon. Thanks for your suggestion 😊

[@eviewei6189]

Thank you so much for this video! I have one question, is it possible to calculate average as the target metric (e.g. avg. sales instead of total sales)? If yes, how do we do that?

[@DataVizCanvas]

Yes you can! You just have to update the cals to average and have to make sure that due to data densification you have to divide it by 2 if it duplicate. Please try and let me know in case of any issues.

[@yuklanng2720]

Thank you so much for your sharing. I was following all steps from your video, however; when I click the category at the dashboard chart, the filter could be match to the category, when I click the filter at dashboard, the chart of the sub-category could not be matched to the category. Would you please advise how to fix it? Thank you.

[@DataVizCanvas]

Hi Yuk, In this tutorial we are using parameter action to assign the category value to the parameter . We then need to use that parameter value in a calculation to filter that category in subcategory sheet. Check out the video at 9:30 where I explained about "creating a dashboard" We have used category parameter and y axis parameter in this tutorial. Also, I have used parameter action to assign the values and we need not to use the category filter to change the value. Hope that clarifies 🙂 Please feel free to reach out if you are trying to use filter for something else and i will try to provide a solution. Thanks Gurpreet

[@yuklanng2720]

@@DataVizCanvas Thanks Gurpreet. Also, I have another question that when I adding the new sheet for "path", all of the data has doubled and it has been affected the data of the other tab as well. All data goes wrong by double even the dashboard of the sales (I have followed the output of your dashboard). Would you please advise how could I fix this issue since I really want to use your chart into my assignment. Do you have the same issue with me? Thank you.

[@NyxView]

Awesome

[@marcellaurentestrada4024]

Thanks for this tutorial it is very helpful! I'd like to show only the top 10 of each level depending on the selection from the previous level. Let's say I select "Furniture", I'd like to show the top 10 sub category from the "Furniture" category. How would I do this? I tried filtering with TC_Rank, it works but it doesn't distribute the lines properly in the chart.

[@juhu5504]

Thank you for sharing!! Can I ask you how did you bring them close to each other so they touch? 16:45 When I bring them together, there's always one on top of the other which create some little white space in between

[@juhu5504]

Also, is it possible to change the Y Parameter (sub-category starting position) using the Category Parameter?

[@DataVizCanvas]

Hi Ju hi, you can achieve this by using floating worksheets rather then tiled one. That way you can align manually and overlap little bit to bring both of them together. In regards to changing the y parameter value, yes you can change based on your requirement as I’m using parameter action to change the value on select. You could do the same if you want to move the start or end position. Hope that answer your question. Please feel free to reach out in case you need more info. Cheers!

[@juhu5504]

​@@DataVizCanvas Thank you for your answers! Regarding 2nd point, I do move the Y parameter on select on category tree but I also want to move it based on Category parameter. Is it possible to use one parameter (Category) to influence the other parameter (Y position)

[@kshitijaddepally6107]

Thank you so much Gurpreet for this video. All your vidoes are very informative. I have some additional queries. I am unable to merge the sheets close to each other even if after using the floating layout. Also I am trying to create 4 levels in my dendogram chart, so when I click on any of the third level branch and when i go back and click on the first level branch, all the previous opened branches from the other levels are still visible. Is there a way by which we can hide the other level branches when we click on the first level???

[@DataVizCanvas]

Hi Kshitij, Thanks for your kind words. You could create any number of levels using this type of chart however i strongly recommend to keep it till 4 levels. In order to create a smooth transition , please check how i used the parameters actions to update the parameter axis values when i select any value. you might need to update multiple parameters when you select particular level. Please try that and let me know in case of any issues. Cheers!

[@pratikpatil-t7u]

Hello Kshitij, please tell me what calculation you have used for y axis in 3rd level

[@mrankit2620]

Hi sir, really appreciate your efforts Is there a way we can have a holistic view of this graph rather than drill down functionality

[@DataVizCanvas]

Hi Ankit, I haven’t tried that with dendrogram chat but it is achievable using Multilevel Sankey chart. Please try that and let me know if you need any assistance. Hope that helps 😊 kzbin.info/www/bejne/fIjVeYxmoNWjors

[@DataVizCanvas]

Hi Ankit, Please refer to below tutorial to see how to create holistic view of dendrogram chart with drill down functionality. Hope that helps 😊 kzbin.info/www/bejne/fIjVeYxmoNWjors

[@mrankit2620]

Thanks a lot

[@aseel6910]

should I make all these steps for every data ? in clear words, when i make a new visualization and I need to make it like this should I repeat all these steps?

[@DataVizCanvas]

Thats correct. I know its not ideal but if you want to create a visualisation this way then you have to repeat these steps using data densification technique.

[@honglindu8599]

Thanks for the tutorial! I have some adjustment questions. I'm not good at the math calculations behind this plot but I wonder 1) how can I extend the length for the start point? Yours are longer than mine. My length before the point (start divergence) is really short. I have checked all formulas and they're the same. 2) How can I adjust the distance between each bar?

[@DataVizCanvas]

Hi, thats a great question. Well the horizontal distance between starting point and end point is determined by X axis value. In this tutorial the range starts from -6 Same way the vertical distance is determined by Y axis calc which will also help to maintain the distance between the bars. Please try and check those calcs and adjust the formula to change it according to your requirement. Please feel free to reach out if you still face any issues. Cheers!

[@igorgolubochanski5485]

Thanks, how I can see the % and not the numbers instead ?

[@olorunfemitunde-adedipe25]

Thanks guys for this. I was able to successfully create a 2-level dendrogram chart. But when I tried to join my charts, so the lines could touch like it is towards the end of the video, it doesn't work for me. The lines of the floating chart tends to overlap on another.

[@DataVizCanvas]

Please double check below list to make sure if everything is aligned: - if you are using floating sheets - if axis on both the sheets are same - if the layout container for floating sheet have same dimensions Please feel free to reach out if you need more information 🙂

[@HooliaV]

This is awesome!! I am having issues with my middle values - for example, the categories that lie in the middle and have a Y axis = 0.0. Clicking on the category with Y axis = 0.0 does not cause any dashboard action to occur (none of the parameters update, and no branches appear.) But the categories with Y axis = 0.1 and -0.1 work fine. Any ideas on what's going wrong? Thank you!!

[@DataVizCanvas]

Hi Hoolia, Seems like there might be some issue with your calculation. When you select the middle category, does your parameter value changes for category along with y axis value? You can see that by showing the parameters. If you can see both the values are changing to desired values then the issue is in the data. Then check if there is a secondary(drill down ) value for the middle category ? If it still doesn't work then check the steps again in the tutorial and make sure you didn't missed any or else share your dashboard public link and have a look and will try to resolve the issue. Tutorial at 9:24 Feel free to reach out if you still have any question / queries .

[@dorotakuleszo451]

hello! Data Viz Canvas... this is awesome video and I managed to do 8 levels using those instructions!! Although I got stuck on the situation where e.g. I am selecting the value from the first chart that updates the second but the rest 6 are not updating...I need to click on each to update. I know that this should be some actions but not sure how to bite it.... any suggestions??? please:)

[@DataVizCanvas]

Hi Dorota, Thanks for your feedback If you have more than 2 levels then in this example you have to select individual values on each level to drill down the next level. Otherwise how you are planning to show the path for each level. There are 2 options you can achieve this: 1. On each level when you select individual category , you need parameter action as already shown in tutorial . Also you need additional calculation which will make the other levels hidden unless you click on the value one the level that you want to be drilled through. 2. You can create an org chart tree structure using below tutorial which might help: kzbin.info/www/bejne/eGGsqXd-baigbpI Please feel free to reach out if you need more info. Gurpreet

[@haileyc7205]

@@DataVizCanvas The videos are super helpful!! In my use case the multi-level dendrogram works better than org chart tree structure because the data is easier to construct (pure SQL, no Python required). Can you please provide more instructions on how to do #1 to hide or show the drill downs in a multi-level dendrogram as in the org tree char?

[@dineshrawat4327]

How to make x and y axis

[@wesleyplaysguitar]

Why are the sales numbers halved? It doesnt reconcile to the data.

[@DataVizCanvas]

Seems like data densification is not implemented correctly. Please check the steps again

[@TheChandansss]

I have an issue overlapping 2 sheets; what should be the floating order?

[@DataVizCanvas]

Hi , you should keep the worksheets next to each other so they are aligned. Please check the video again at around 16:32 on how i have aligned the worksheets. hope that helps :-)

[@TheChandansss]

@@DataVizCanvas Thanks.

[@athirab6183]

I have followed the same steps and created a beautiful dendogram! Everything is working as expected. Only issue is the points of one level are not touching the next chart. My charts are in floating layout. My sub categories change when I click each categories. My containers are aligned. My axis are aligned. Only issue is the points not touching. Anyone know how to fix that?

[@DataVizCanvas]

Could you please check if axis on both charts are fixed and have the same range. If not then once you make it fixed, it will be aligned. Hope that helps

[@ColinMcDuffie]

What if you have many levels to the hierarchy and building the multi-level causes the tableau pop up to tell you have too many rows

[@DataVizCanvas]

Well in that case this visualisation is not suitable for this use case. You need to filter out the data or check out my another tutorial on org chart. kzbin.info/www/bejne/eGGsqXd-baigbpI

[@sconnell194]

👍

[@adorableanimals5559]

How much experience do you have in tableau ? Is this the level required to become a tableau developer?

[@DataVizCanvas]

That’s a great question. For tableau developer role you need to have basic knowledge of tableau and the creative skill to tell the data story through your visualisation. You also need to know SQL and few other skills like requirements gathering. Talking to stakeholder to understand business problem etc. The advanced chart in this tutorial is not required in tableau developer role buts it’s added bonus. Hope that helps

[@pratikpatil-t7u]

Thanks for the wonderful training but in my visualization I need total 4 level while moving to next level Y axis calculation is not coming correctly all the values coming from either -ve side or + ve side Formula I used as below, please confirm if its right or not. [@sigmoid]*([@Rank]-(WINDOW_MAX([@Rank])+1)/2)/100 + [@Y Axis Para AndonType] + [@Y Axis Para ReasonGrp]

[@mohammadzaman7383]

Thank you !

[@DataVizCanvas]

You're welcome!