R, Visualization

What’s new in Tableau 8.1?

We had a great customer conference in Washington, DC. Many new 8.1 and 8.2 features were unveiled including the long awaited Tableau for Mac. My teams built a significant number of new features coming in the 8.1 release coming this fall and here are the highlights :

R integration : Always on the cutting edge of data analysis? In Tableau 8.1 you can author calculated fields using R which brings a vast array of statistical libraries to your fingertips!

Box-Whisker Plots : In Tableau 8.1, you can create Box-Whisker plots with a single click. We offer a number of preset styles, color options, plot types (Skeletal or Schematic box plot) and ability to show/hide marks.

Ranking : Adding rank is now a one click operation and how you want to treat the ties is entirely up to you. The options are Competition, Modified Competition, Dense Rank, Percentile Rank and Ordinal Rank (Unique).

Percentile Aggregation : It has been possible to get 50th Percentile using Median aggregation but now you can use the new aggregation Percentile to get 5, 10, 25, 50, 75, 90 or 95th percentile by directly interacting with the pill or any percentile of your choice using the Percentile function in a calculation.

Two-pass totals : Prior to 8.1, Tableau Grand Totals were calculated by applying the same aggregation over the entire dataset. Calculating something like daily average sales from transaction level data (AVG of SUM(sales for a given day)) was not possible using grand and/or subtotals and required a table calculation. Now it takes a single click to do the same! You can specify the type of aggregation and whether or not a total value should be shown per measure.

Custom Start of Week :  If you are in the EU, now you can use Monday as your start of week and all the date calculations (DATEDIFF, DATEPART, DATETRUNC…), as well as aggregations respect this setting. If your company or country uses a different day of week, no problem. You can set it to any day! You can also store this information in published data sources making it easy to ensure that all users have the correct standard setting used across the company.

Forecasting Improvements : Forecasting got a whole lot better with addition of support for multiplicative models and prediction bands. Now it works also with measure names/measure values, multiple measures and fiscal calendar.

Stored Procedure Support : Tableau 8.1 supports interacting with Sybase ASE, Teradata and SQL Server stored procedures like accessing tables. You can select the stored procedure just like picking a table in the connection dialog and either type in values for parameters or use Tableau parameters so you can allow users to tweak the inputs to the stored procedures using widgets in your visualization.

SAP BW Aliases : In Tableau 8.1, you can right click on a dimension and pick the alias you’d like to see in your visualization. Key, short name, medium name, long name and their combinations (separated by a dash) are supported.

SAP BW Extracts : Now you can extract data from SAP BW cubes into Tableau’s data engine as well. Tableau retrieves leaf level nodes (not drill-through level data) and makes them into a relational data source. Since multidimensional to relational transformation does not preserve all cube structures, switching back and forth between extract and live connection freely without impacting the state of your visualization is not supported for cube extracts. You will need to make your choice before you start building up your viz. However you don’t have to decide on everything upfront. You can switch between alias options (key, long name etc.) after the extraction.

Join culling : If you have a lot of joins but no referential integrity defined in the database for performance reasons, you are going to love this. Now you can impose Soft RI in Tableau by simply checking a box and Tableau will eliminate unnecessary INNER JOINs, making your queries run faster.

Initial SQL for Aster Data : Starting with 8.1 you can write initial SQL for Aster which will be especially useful for SQL MR functions that require output tables as inputs to later stages of the analysis (e.g. cfilter) and cannot be expressed with nesting in a single query.

Google Analytics Advanced Segments : Tired of down-sampled data because you ended up downloading the entire dataset? Now you can use both default and advanced/custom segments in Google Analytics. Interested in search engine traffic, mobile traffic… just pick what you need.

Google Big Query BQL Support : Say hi to big JOINs and GROUP BYs… Starting in 8.1,  you can use BQL in Tableau and take advantage of special constructs that are not supported in SQL 92. You will also notice a number of BigQuery functions appearing in Tableau’s list of calculations when creating your calculated fields that will come in handy for parsing URLs, concatenating strings, converting Unix timestamps to dates (and vice versa)  and when you need to use regular expressions.

Teradata OLAP Connector : Tableau 8.1 has a  new first class connector for Teradata OLAP with named sets and drill-through support.

Cross-version Compatibility between Server and Desktop: Using a Tableau 8.1 desktop, now you can discover and issue queries against data sources published on a Tableau 8.0 server.

Not to mention over a dozen query performance improvements that will make working with live database connections smoother. 

Backstage of TCC 2013 keynote

DATA TO THE PEOPLE : Things were hectic in the backstage for the TCC 2013 keynote

This is just the list delivered by my teams. Tableau 8.1 is packed with many  more features. Grab a copy of the Beta and share your feedback with us!

And of course… Stay tuned for tutorials on the new features.


18 thoughts on “What’s new in Tableau 8.1?

  1. Sid says:

    Hey Bora

    Could you expand a bit more about these whenever you find time? I am a newbie and looking to reap the benefits of this upgrade as much as possible.

    1. Join culling – If you have a lot of joins but no referential integrity – What do you mean by this. Can you give quick example?

    2. Two-pass totals – Can I show something like (sum(max(Lines))) ? What if I have 2 hierarchies – Order and Order line – For Order line I show just max(lines) and for Order I show sum(max(lines)) when user changes hierarchy? Will this be possible

    3. Forecasting Improvements – “Now it works also with measure names/measure values, multiple measures and fiscal calendar.” = Can you expand on this ?

    4. Ranking – What are these different kinds of rankings?

    I am just starting into the world of Business Intelligence and hence your explanation would mean a lot.


    • Hi Sid,
      1.Let’s say you have Price from TableA and Product from TableB and you joined these two tables. If you just have AVG(Price) in your visualization, Tableau can get you that information by querying TableA only (so it ignores the join). Before 8.1, this required primary-foreign key relationships defined in the database. In 8.1 if your DB admin did not define primary-foreign key relationships for performance reasons, in Tableau you can tell there is such a relationship and Tableau will behave the same way it would if they were defined in the DB. This only works with INNER joins.
      2. Yes. It is adjustable per pill. So if you have a column sales and another one profit. Inside the same viz you can have one of them display as AVG(SUM(Sales)) and another one MAX(SUM(Sales)) etc.
      3. Basically you can have multiple forecasts inside the same visualization. Prior to 8.1, you could only have one. Also before 8.1, forecast option would be disabled if your date axis was fiscal calendar. Now it works regardless.
      4. Different ranking options treat ties differently e.g. if you have two people sharing the 2nd place, one of them would rank people as 1,2,2,3, another option would rank it as 1,3,3,4, percentile rank would give you the results in percentiles like converting test scores to percentiles…

      I hope this helps.

  2. Sujitha says:

    Hello Bora,

    I am a newbie to Tableau and I would like to connect it to the SQL Stored Procedure. I saw a couple of articles on how to connect to a SP, but I am not sure how to pass the input parameters dynamically. Could you please let me know if you have any idea?

    • Hi Sujitha,
      Did you try using Tableau parameters? Instead of typing in a value you can insert a parameter (you should see the option in the stored procedure dialog). Then you can link that parameter with a text box or a drop down with prepopulated items.

      I hope this helps.

      ~ Bora

  3. Hi Bora,
    I reached out to my Tableau sales rep about obtaining a different product key for the SAP BW data extract feature in 8.1 and she said that we would probably have to contact SAP about this issue. How/who are we supposed to contact for this key?

    Thanks in advance!

  4. pavan says:

    I want to replace Null values with N/A,can you please help me in this,as far as i know we need to create a calculated field using IFNULL,.what is the exact command?
    Thanks for above quick reply.

    • Since you’re trying to convert to n.a, I assume you mean before sending to R. Tableau already passes NULLs to R as N/A. If you have nulls in your data, SCRIPT_BOOL(‘is.na(.arg1)’,ATTR([Your field here])) should return true for the rows that have null. We had a bug in 8.1 where NULLs were passed as 9.65404271973796e-321 but this was fixed in 8.1.3 maintenance release.

      • pavan says:

        sorry for late reply.
        i want to show null values as n/a in my view to users.
        that is my requirement

      • Depends on your view. You can go to Format > Pane > Special Values and enter what you want in there. If you want to do it in a calculated field IFNULL([Country],’n/a’)

      • pavan says:

        sir,actually i am new to tableau,recently i attended an intervew.
        so,i dont have answers or doubts on some questions.
        i am presenting them here.
        Can you help in this please?

        1)What is assume referential integraty,what is the use of that?
        2)what is default data blending join what if i want to change it?
        3)max no.of records in a table?
        4)difference between scatter plot vs tree map?
        5)types of filters?
        6)types of annotations?

        it will be a great help to me if you reply as i have another interview tomorrow?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s