Where is the column header date format conrols?

Hey, hive mind! :slight_smile:

I want to change the date format for the “month, day, year” portion of the header in this metric set:

I’ve looked everywhere I could think of to find the menu that lets me change the date format. The rest of my team has had a look as well.

What are we missing?

Do you want to change it to something like 2021-09-15 ?

If so, I think you’d have to do it at the cube level. I’ve run into the same thing, as I want to do custom text for row headers but can’t.

I do also see Custom Text down there - have you tried playing with that? Often you can do [var:yyyy-MM-dd] (where var is your date column’s name) and get it to work (IF it is a time dimension).

Ken - thanks for the suggestion!

We eventually got where we wanted to go by adding in different versions of the date field. One trick to that was that we had to order them in the data analysis panel (vs. the visualization panel) to get them to show in the desired order on the table:

image

The result gave what we wanted:

With monthly totals, weekly subtotals, and the day of the week showing for each column.

Another puzzle solved! Though I still wonder if I could have done it the way you suggest…

1 Like

Hi Thomas,

This is a nice way to break down dates without repeating all the parts of the date in each column. I can respond to the original question though:

I would say the primary place to change the formatting of dates & times is in a time dimension. In this case you already seem to be using one, and if you edit it there is a formatting section you can expand to customize each time dimension hierarchy. This would apply everywhere that hierarchy is used, although you can create and use multiple time dimensions and choose between them: https://www.dundas.com/support/learning/documentation/analyze-data/create-a-time-dimension#format

Tables can also override the formatting in cases like a regular column of dates like Ken mentioned, but these options were not really meant to replace or duplicate time dimension formatting as long as those options work for you.

2 Likes

Jamie,

Thanks, but I find I’m not able to follow the instructions in the link that you sent.

https://www.dundas.com/support/learning/documentation/analyze-data/create-a-time-dimension#format

The instructions tell me that I can format the time dimension’s levels “when you open an existing time dimension from the main menu.” I’ve tried all combinations that I could think of, trying to open the dimension from the data analysis panel, visualization panel, explore window…left click, right click. I don’t see the menu that is being shown in the instructions.

Instead I see this:


I don’t find a “date range” or a “formatting” section.

What am I doing wrong?

Hi Thomas,

What we call the ‘main menu’ is the black menu along the left, where you can go to Open > Time Dimension. The beginning of that article shows creating a new time dimension from the main menu clicking New instead of Open. In the Explore window, you can also right-click a time dimension found under the “Time Dimensions” folder to edit it. Then you can set Hierarchy Level to Day to customize the format for the Day level for when you have a table like the one shown in your first post.

This is for customizing the formatting of a time dimension hierarchy such as Year > Month > Day, but your screenshot shows the dialog accessed from the Data Analysis Panel for customizing the Day Of Week attribute, which is sort of the alternative you found later to instead drag each component of the date separately (Month of Year, Day of Month, etc.)

Jamie,

Thanks, I completely misunderstood the article. I’d skipped ahead to the “Formatting Time Dimension” part and didn’t realize that the article was only talking about time dimensions that are created separately from my data cube. I was trying to change the appearance of a dimension containing date data.

Is there a different term to use than “time dimension” for dimensions in a data cube that contain time data, to keep it separate from time dimensions that are created separately from the data cube?