[Tip 1 - 39] Dundas BI Tips & Tricks!

Very useful, how can I return to that page after completing this challenge?

Tip 32 - Ensuring Data Integrity


To correct cases where you have different categorical values in the data that should actually represent a single value (i.e.USA/United States/United States of America/ etc.), you can easily apply correction using the string transform in the data cube layer.


For example, if you wanted to change all the "United States" values in your data to "USA", simply add a string transform to your data cube, choose the replace function and set "United States" as your old value and "USA" as your new value and you're done - your data is now clean and consistent without any scripting.


To learn more, check out: http://www.dundas.com/support/learning/documentation/data-metrics/transforms/common/string



Thanks

Tip 33 - Creating a Custom Interactive User Experience Without Scripting


Did you know that you can easily setup interactions not only on top of data visualizations but also on top of components (non-data controls such as drop-down lists/radio buttons/labels/etc.)? This allows you to easily create a unique user experience that goes beyond standard drill-downs and filters.


For example: you can create a custom navigation experience based on a user selection from a drop-down list - all without scripting - simply by binding the interaction to the select drop-down list value element.


For exact steps and other options - please follow: http://www.dundas.com/support/learning/documentation/design-view/interactions/using-interactions-with-controls


Tip 34 - How to Handle Missing (Null) Data


Often times when we group data by certain hierarchies, some of the data points are missing, which results in broken trends or incomplete data sets. A quick fix for this issue - without having to go back to the data source - is to define a rule for it under Dundas BI and have the system complete the missing data for you. The rules you set can be to assign missing data points the value of zero, or the average across the entire data set, etc.


You can read more about the different rules, as well as how to set it up here: https://www.dundas.com/support/support-center/support-articles/data-visualizations/handling-null-data-points

Tip 35 - Convert Measures to Hierarches on the Fly


When exploring data directly from a data connector, you will notice that Dundas BI attempts to automatically identify measures and dimensions in your data. Measures will be marked in green and dimensions will be in orange. In some cases, the detection may be different than the setup you would want, to better explore the data.


For example - if the field in your data is "management levels" and the values are all numeric (1, 2, 3, etc.), it may be detected as a measure, while you may want to analyze it as a dimension (i.e. group all employees by management levels to see the count under each level). In order to treat this field as a dimension - all you need to do is right click it and use the "Change Category" menu to switch it from a measure to a dimension. Then you will notice the field is colored in orange, and you can drag it into your rows or columns section of the data analysis panel to group by that data field

Tip 36 - Quick Statistics for Rapid Analysis


Did you know that Dundas BI allows you to select data points on your visual and get quick statistics such as sum, average, count, etc. on the summary bar?


If you love using this option in Excel, you would love to use it in Dundas BI! And not only on tables, but on any visual you use to represent your data!


http://www.dundas.com/files/theCube/Summary.png

Tip 37 - Watermarking Sanctioned/Non-Sanctioned Data


Often times, a requirement that is associated with better data governance is the ability to watermark content based on approved/non-approved data. This is most important if some of your power users are conducting their own self-service analysis with IT certified data as well as personal data files, and end up sharing those with others.


You're in luck! Dundas BI provides the ability to watermark data sets and content based on whether or not sanctioned data sources are used!


http://www.dundas.com/files/theCube/Watermarks.png


Dashboard with watermarks in Dundas BI


The only prerequisites you need to accomplish this, is basic familiarity with JavaScript and CSS.


For more information, go here: http://www.dundas.com/support/support-center/support-articles/scripting/watermarking-sanctioned/non-sanctioned-data

Tip 38 - Job Failure Email


With the release of Dundas BI 4, you can now configure Dundas BI to send email notifications to specific accounts when a job fails. For example, a data cube storage to the in-memory engine failure. This configuration option can be found under:


https://www.dundas.com/support/secondary/view-or-modify-configuration-settings?v=4.0#h4-new-or-updated-settings


If this is of interest to you, be sure to set it up!

Used this tip in a couple of dashboards

Used this in a line chart legend and worked perfectly however it has its limitations in other chart types

Used this in several dashboard

Great forum and smart ideas. ..

Tip 39 - View the Queries Sent to your Database


Starting with Dundas BI 4, you can now change your log filter settings to see the queries sent back to your database. This can be very helpful in case you want to troubleshoot a specific scenario using the query created by the system. You can learn more about this feature:


HERE

use this tip all the time!

Hello Jordan, is the code or the dashboard available for the second option, I prefer the interactions of that one and would like to apply it to my dashboards.

Hey Doug,


If you’re interested in applying those interactions to your dashboard, you don’t actually need code for it. If you follow the steps above in the 3rd paragraph, and use an image, rather than a button, you should be able to accomplish this.


A more detailed example of how this can be done, can be found here: http://www.dundas.com/support/learning/documentation/design-view/interactions/use-a-button-to-apply-a-filter-value

Hello Jordan,

I have looked at that, and it does not have the iteractivity between multiple buttons. In the second example when you click the first button it becomes highlighted as selected (something I cannot seem to get the new filter buttons to do). But more importantly, when you click a different button, not only does the filter change but the state of the first button becomes unselected, in otherwords they are responsive to each other. Additionally, I am hoping the code will help me learn to make a button two-state, so that when I click on it once it is selected, when I click on it again, it is not selected. two actions for click dependent on current state. I hope this makes sense.

Hey Doug,

You are right, the interaction itself will not highlight the button or the image. What you are looking for is a styling code, which can be written in countless ways. Basically, you apply one style to the control that was just clicked, and another to all the other controls.


In this specific sample, we used updateOpacity and set it to true for the clicked group of controls. For example, the Asia group has this script on the click action:


hospitality.updateOpacity("AsiaGroup", true);
hospitality.updateOpacity("NorthAmericaGroup");
hospitality.updateOpacity("EuropeGroup");
hospitality.updateOpacity("OtherGroup");

I am prompted for an ID and password, what should I be using there?