How to create visualization with non numeric columns only

[former_member737689]

I have a dataset which looks like this. (see attached png)

Cast Genre

actor1 Kids

actor2 Comedy

actor3 Kids

actor1 Comedy

How can I create visualization using these? Let's say I want to get a bar chart of actors (cast) and count of shows they have acted for every Genre. How can one do it? since there are no fields that are measures.

[DebjitSingha · Accepted solution]

Hello shravya.kilar,

I have see this kind of requirements commonly in M&E industries were some time reports are requested based on MPM's.
What you can do it create a counter field at the transformation level. This way this counter will always gets updated whenever you reload/ refresh SAC model. Here is an example, you can choose any form of combination and formula based on multiple conditions. -

Create model based on dim fields.

Add a formal filed from top. Then based on your requirement you can choose to pick multiple fields for validation. Here I just select Cast columns and checking if it is not null. Else create an entry with "1".

Convert the newly created filed "counter" as measure.

Create model. You can test the transformation (defined above) by uploading another set of data file (from data management tab or connecting model to import connection)

There are multiple ways to address this requirement, though I found creating a counter at the model level is the most useful.

Hope this helps. Let us know if you have any query around this topic.

Viewers/ readers - If you find above information helpful, feel free to up-vote (arrows on left side).

- Debjit

[former_member751560 · Accepted solution]

Is it also possible to show the counts instead of the percentage?

In absolute numbers, that would be very useful in many plots.

[DebjitSingha · Accepted solution]

tqbf yes you can.

[Former Member]

Any graph of non-numerical data must show two things: the categories and their sizes. The two most popular forms are the bar graph and the pie chart, which you're probably already familiar with.

Raw counts, known as frequencies, or percentages, known as relative frequencies, may be used to represent the sizes of groups. (Relative frequencies can also be expressed as decimals, but I believe most people understand “20 percent” better than “.20.”) What considerations do you weigh when determining whether to view frequencies or relative frequencies? This is a personal preference, not a matter of right or wrong. Your decision is based on what's important to you and what argument you're trying to make. If your main concern is with the individuals in your study, frequencies are the way to go. However, if you want to demonstrate the parts' relationship to the whole, use relative frequencies.

Hope this helps.

If you want further info, plz contact me on the address on the given website

[LINK REMOVED BY MODERATOR]