trafalger Coach

You could append two datasets together and switch back and forth with a filter or a variable - but I agree with @MichelleH, more information is needed.

Thank you @GrantStowell, @GrantSmith, @marcel_luthi and @Ashleigh!

Would something like this work? Case when SUM(SalesQuote) <> 0 and SUM(BannerAverage) <> 0 then 'Show' else 'Hide' end

With out of the box Domo functionality, I don't think this is possible. It might be possible with campaigns, but I've had the same issue and just end up sending 4 emails or if it makes sense combing the datasets into 1 and having the card show that.

Not sure if you could do this, but there are probably easier solutions. You could make a beast mode that says if it's the first of the month, show the prior month, else show the current month.

So do the same logic, but use @GrantSmith's post here on how to calculate rolling averages:

Just something like this:

Great job @DavidChurchman!

How are you calculating the average? Could you just go a group-by in the ETL by Creative type and then append it to the dataset? Then if anything new is added it would just get added automatically.

Thanks @DDalt!

I haven't experienced it in the handful of environments I was in this morning

Have you tried the grouped bar + line chart type? Your data does need to be structured a certain way to get that to work EDIT: I meant grouped, like @Juliann_Potter recommended.

I've done this before but by concating fields from Domo into the URL of the system, like companyid.netsuite.com/customer/DOMOFIELD. Those specific systems had the IDs of the things in the URL so it was just a matter of recreating the URL syntax in Domo

For previous year to date, I do: "Case when Date >= DATE_SUB(CurrentDate, interval 1 year) and Year(Date) = Year(DATE_SUB(CurrentDate, interval 1 year) ) then Value End" Find the year value of today's date 1 year ago and only selects dates less than or equal to today 1 year ago. 

I believe it will keep 9/01 open and then also have 9/08 open. I submitted an Ideas Exchange topic about auto-collapsing:

For previous year to date, I do: "Case when Date >= DATE_SUB(CurrentDate, interval 1 year) and Year(Date) = Year(DATE_SUB(CurrentDate, interval 1 year) ) then Value End" Find the year value of today's date 1 year ago and only selects dates less than or equal to today 1 year ago.

Agreed, I ran into this issue yesterday and was frustrated I couldn't do this. Great request @marcel_luthi!

Why are you using the LAST_DAY beast mode here? That will make all of the dates the last day of the month that they're in.

It's irritating that NULL values tend to break math

Do you need the card to respond to page filters/date range? Because you could sum up the ChildPN's active qty in the dataflow and then join it back on and then filter on that? (Which based on your question, I think you do). Could you remove the line items from this view, filter the ChildPN here and then have a drill that…

If I remember correctly you'll need to link pages to cards to datasets, which can cause some cardinality issues, but then you can also credit a card/dataset when a page gets viewed vs. just when a card gets viewed.

Agreed with the others, I don't think you need the FIXED portion, as just a count(distinct) should filter it out. I had a client who did something similar and it worked great.

Agreed, I've done dynamic column renaming using a Stored Proc in a MySQL dataflow - you can query the system tables to get the schema.

What does the card look like? Do you need the partition? If the card is using ChildPN on it you can just do a sum and then then filter on that.

Actually, no that's wrong. There's already an IN in pfilters. I'm not sure how contains functions.

The documentation is here: If I remember correctly, it functions similar to an "in." So contains ("ItemA", "ItemB", "ItemC") means if it's Item A, B or C it will filter.

I know some Bar graphs do, but it's only like "most recent month's value"

Are any actions captured in the activity log?

In this case I'd join to the date dimension, find all of the days between TimeOffDate and ReturnToWork and split the HoursOff between them. Similar to this:

I'm not sure on this - I'd try reaching out to support and see if they can assist.