Best Of
Re: Number of week calculations for fiscal YYWW (Year, Week of year)
Thank you @GrantSmith - Your advice is what solved this for me.
As suggested by GrantSmith, I created an excel upload dataset with a file that references our 4 digit YYWW to every calendar date in that week. Here is a sample of the spreadsheet used to create the date conversion dataset:
The dataset was created by uploading the excel to DOMO:
The card I was creating was already using a Magic ETL dataset:
At the bottom, you can see where the date conversion dataset is coming in to the existing dataset that was already almost what was needed. DOMO didn’t automatically identify the MMDDYYYY fields coming in as dates, so I put in the “Alter Columns” module to change the properties of these fields to an actual date data type:
Since there were 2 YYWW fields to be compared, added 2 joins. Each join was configured similarly to associate a YYWW field from the main data to the newly brought in MMDDYYY date formatted data:
Important note: With the second join there was a warning that there were duplicate field names, which forced renaming them to something different on the 2nd join. I chose to name the field in a way to reference to the respective column to make it easier to tell the difference later.
Used a Select columns module to rename and move the new columns to a place that made them easy to use later and rename them:
Looking at the dataset, there are now 2 places where there is a MMDDYYY date field which correctly corresponds to the YYWW date in the neighboring column – Note, I chose the Sunday day of the week on both as our systems update on Sunday making it the most relevant date to choose – any date would have worked though as we are only trying to get a count of the difference in the number of weeks.
Which now makes the beast mode super easy to write for the card:
And gives a result we can display and use for filtering:
Thanks again for pointing me in the right direction, Grant!
Brent Jensen
Bass Pro Shops
Merchandise Operations Analyst
Re: Filtering Report Data by Multiple Dates
@Brightsider you can but you would have to restructure your dataset. You'd need a single date column to filter off of, so you'd want to pivot your data to create a single date column, and then have a new label column that holds the value of "Created" or "Scheduled"
mhouston
Re: Export All Names of Beast Mode Calculations
@newbee do you have dataset access? You can get this from the Domo Governance connector using the "Beast Modes"report - then you could create a view on top to filter by the dataset you are concerned with.
mhouston
Re: How to Make a Weekly date range into a Daily Date?
@WorldWarHulk I have done this with similar data from Safegraph. You should be able to do this in a Magic ETL. First step is to use the Unpivot command. To get the data to be something like thins:
Start Week, End Week, Dayof Week, Value
With Day of Week being the column name Sunday to Saturday.
Then you can do a formula to get the actual date. Something like this
case
when DayOfWeek='Monday' then dateadd(startweek,1)
when DayOfWeek='Tuesday' then dateadd(startweek,1)
You will need to check the formula syntax in the formula editor in magic. but hopefully you get the idea here.
Re: Change a Dataset and Replace All Downstream Processes With It?
I feel like there might be a way to force the dataset ID to match the ID that the WB was using. If you did that and deleted the WB dataset, would that work? Kind of theory crafting here... but I thought the other ETLs, etc. pulled the dataset in by looking at the dataset ID. So, as long as your replacement dataset had the same dataset ID, wouldn't that work?












