how do you know which row you want to exclude?

i wonder if you could get that from the layouts API. just edit your dashboard, right click and enable monitor network traffic in your browser, then save your dashboard, search for layouts in the resulting list of API calls, I'll be the type of interaction is captured in there somewhere.

you could materialize your FIXED function as a column on your dataset by implementing it in DatasetViews. It seems like you need MAX(MAX() ) because you're trying to ascertain "current_week_num" , that's static and probably doesn't have to respond to filter criteria, so just add it as a column.

@Jmoreno you can install workbench on your desktop

@klifford make a date dimension table that has one row per day and a column for all the fields you'd want to pre-aggregate on. @GrantSmith did a nice writeup here https://dojo.domo.com/main/discussion/53481/a-more-flexible-way-to-do-period-over-period-comparisons#latest

@MarkSnodgrass i was flipping through your tutorial. in your proposed solution the DateFilter is set to This Year + a couple extra years via YOY comparison option. as best you can tell, is Domo actually applying the DateFilter filter? I'm guessing not b/c Domo ought to be limiting you to either just this year, or the last…

surely this is a bug. not an intentional design choice.

This request has been a known missing feature in Domo for years. It would be amazing to see headway on this.

Some videos on Domo Everywhere from my conference and channel. https://www.youtube.com/watch?v=jcJ3xSCzbdQ

Imagine you had C-Suite , VP, Director | 19 | Enterprise, Supply Chain You'd need to duplicate the row a total of 6 times. take the first column and duplicate each row by the number of Commas +1 (because that's how many respondents there are In this example ,you'd want to duplicate that row 3 times. then use split-part()…

i'm not sure if i understand what you're trying to accomplish. it sounds like you're trying to create a sum across rows. that's easy with a formula or beast mode (col1+col2+col3+col4). if you want to use the past n rows, you might look at using the LAG() function in the RANK and WINDOW tile. it's a google-able operation…

You could look at customizing user experience with different arrays of feature switches in multiple instances of Domo. if domo ever gets an instance switching feature kind of like Buzz on Desktop, it could create a seamless experience for having different views with different feature switches. either way this sounds like a…

@Mychele you want ... everyone to be able to edit the beast mode? or you just want them to see the definition? b/c you can accomplish "seeing the calculation" with domo_governance + BeastModes report. if you were building a card, that could be accomplished with pfilters.

+100 It'd be a great sanity test to know, "no you're not crazy", especially when Domo makes silent product updates that have unexpected impact on other areas of the platform.

it's not in product yet, but you can do this with scripting. https://www.youtube.com/watch?v=v-8HYp6b4KU

i hear what you're saying, but tagging a dataset is the same number of clicks as moving something into a different folder... in many ways tagging structures are more powerful than folders because a dataset could be attributed to multiple tags (whereas a dataset couldn't live in multiple folders). why not just use Tags +…

build a fiscal calendar where for each day you assign the week the date belongs to. if you can imagine, every company would have a slightly different take on what the week number should be, so having an artifact in a table 1) guarantees that you can see it 2) creates a column you can filter any dataset with.

if you wanted to just export the transform steps you can do that VIA API. since just knowing the transform steps can be obtuse, typically we'll recommend users generate documentation in lucid chart or similar to document the dataflow. why would you give such an intimate view of the ETL without also giving edit rights?

you could build a table card from domo_governance_pages report that accumulates history and captures last_updated_date over time. So if a card didn't previously exist or was updated since the last report run, you could build in logic for a flag and how it gets displayed / alerted in the card. but yes... it's a shame that…

@louiswatson you could have significantly simplified your ETL by using a FORMULA tile with a CASE statement. The reason why this approach is cleaner is the CASE statement allows you to use an ELSE clause. When you split out your data using FILTER tiles as you have, if you miss a new filter value you exclude it from your…

@Jose_Rojas yes, you would need to work with your AE to get the Jupyter Notebook addon. If it was me, i would be VERY explicit and adamant that you do NOT need the Data Science Package. IMHO with the improvements in the Jupyter Data Science Notebook there is no reason to use the Scripting tiles. But also i think you're…

@utkarsh please articulate in business language what you are trying to accomplish. @MichelleH is correct, you can probably accomplish what you need using Window functions (the LAG() function) in particular, but to assess whether it would always be appropriate to use LAG it's important to understand your use case.

build a date dimension where you have one row for each day of the year. then assign the week_of_year to each date. then it'll be easier to make these comparisons. "build a date dimension" in SQL is a google-able term, so you should be able to implement that using MySQL dataflows.

@julia_wang it would be helpful to know what columns you have on the axis of your pivot table. (post a screen shot). also please post the beast mode you've written and include data structure... makes it easier to guess where you're going wrong.

@AdamC Domo predominately uses MySQL syntax.

if this is a bug, probably should open a support ticket. if it's a product feedback, make sure to put it under the IDEAS exchange board. if you leave it here i am concerned you won't get a response.

@chapman just write a CASE statement. SUM(SUM(case when company = 'abc' then `totalSales` end ) FIXED (FILTER DENY company))

Maybe an "include / exclude current_date" option.

consider using window functions to get the desired result. if you UNION your data (Consumption UNION to Inventory with a column activity_type = "consumption" or "inventory" so you can tell them apart) as @MarkSnodgrass recommended and structure your data such that you have Activity_Date and Activity_YearMonth, you could…

@mHejl if it were me i would unpivot the data (Magic ETL) to collapse the values in columns 2, 3, and 4 into one column then do a windowed MIN function with the Rank and Window tile. If you don't want to restructure your data, you could see if Magic or Beast Modes have an implementation of the LEAST() function…