ST_Superman Domo Employee

I'm not sure that this is what @hamza_123 is after here. It sounds to me like he wants to create a card that will provide the name of the student that received the highest score. I suppose if you sorted the card by max(`Score`) then you would see the student name in the first bar or line of data. You could even then click…

you could also use gauge cards and the story dashboard to provide multiple summaries...

OK. So this is what I did. I'm not sure what your dataset looks like, but I assume you have a Date, Employee (either ID or name), and a field for hours worked each day. I made a dataset with dates from 4/1/2019 to 6/30/2019 (Q2) and populated it with 5 users and a random number between 0 and 8 for each day. Then I used…

Are there any null values for the date field? What are the names that are missing? It’s also possible that if your trim() or concat() functions are erring out then it won’t count either. 

For now.... you can upvote here https://dojo.domo.com/t5/Ideas-Exchange/Magic-ETL-Commentary/idi-p/26073

How many days are in your work week? Assuming 5 days, then you could calculate utilization using something like this: sum(`Hours`) / (32 * CEILING(COUNT(DISTINCT `Date`) / 5)) This would look at the number of dates that a user has entered hours for and divide it by 5. If they logged more than 5 days, say 7 days... then it…

If you use the group by tile to find the earliest test start date per `test taker: Resource Name` and then join that back to the original data, you can build your graph.

Is there a reason that you don't use the five9 account number? Why are we counting unique names?

Yes, you would want to calculate the earliest start date for each user. Then you could create the graph using the earliest date as the x-axis and the distinct count of resource name as the y-axis.

Not if you are plotting it by month. The summary number, if you are using the distinct count funciton in your first post, should still give the correct overall number of distinct names. To demonstrate, I duplicated the 5 rows of data you provided twice and changed the month to be Feb and March. Here is the output. You can…

In this case... remove the MIN sorting. You will still end up with duplicates anytime the same user takes a test in multiple months. The calculation will be counting the number of unique users that started a test in a certain month. The overall summary number should still be accurate though.

Also, it looks like when you drill in, you are using "corrected test date" in your table. This could be part of the issue. You changed your graph type on me from your initial post ?

I'm not sure why you are getting a different result

Can you take a screenshot of your entire analyzer screen? it may help to see what else might be going on. It may also help to get a few sample rows of data with the same schema as your data set to help troubleshoot.

The ODBC connection could be used to pull into any environment that accepts an ODBC connection. You can pull it into r studio, python, MySQL, redshift... possibilities are endless.

This is AMAZING! How did I miss this in the release notes? Thank you @takap for bringing this to my attention. For those of you that don't want to copy and paste into google translate... takap has pointed out that Dataset schema information description column is now displayed in analyzer mode

@user19690 Can you provide a few sample rows of your data set? My guess is that something like this should work: ( SUM(`headcount`) - SUM(`leavers`) ) / SUM(`headcount`) But that would depend entirely on how your dataset is built

You can get access to a DOMO ODBC connector which will let you access your Domo datasets directly in your MySQL workbench environment. This is currently a premium product and has an associated cost. You should contact your Domo Customer Success Manager for more details.

What defines a "unique blood draw"? Presumably there is some kind of draw_id field? In that case, you would want to do something like this in your beastmode: COUNT(DISTINCT CASE WHEN `Successful/Unsuccessful` = 'Successful' then `draw_id` END ) / COUNT(DISTINCT `draw_id` )

Adding on to @Godiepi ... If you want a distinct count and need to display the first date the name was in the data, then change the agregate on the first column (the date column) to be MIN If you are more interested in the most recent date, then use MAX. The reason you are seeing multiple "distinct" names is because the…

I think you will need to calculate your YoY and QoQ values in a dataflow. Let me know if you prefer to use MySQL dataflows or ETL and I can try to walk you through the steps. Beastmode calculations are not able to look at multiple rows at a time (unless you are aggregating them, but then it can only look at one aggregate…

You should be able to sort by just placing the date field in the sort section. Even if you are using the beast mode calculation as the label. 

Are you using the Pivot tables? You should be able to use the actual date field as your Columns and then select "graph by" Month. I'm not sure about your second question. I don't think there is a way to hide those labels.

I would use three tiles to do this in ETL. First you need to create a constant field that will contain the current date. Then you perform a date operation to find the month of your date field (called `Date` in my dataset) and the month of the Current Date that was just added. Finally, you can filter your dataset to only…

You can do this in a MySQL dataflow by using this: SELECT DATE_FORMAT(`Date_Field`, "%M - %Y") as `Month - Year` FROM your_table

Tell me if I’m understanding this correctly. You have three columns with dates in them and you need to keep all of the data but have only one date field?

The only way that I can think to accomplish this would be through some sneaky data engineering. You would need to be careful about how the dataset was going to be used because anyone not familiar with the structure would likely double or triple count you value. That being said, what I think you want to do is to "stack"…

I see two ways to display this data. I'll try to explain both so you can pick which you prefer. The first option is to create a flex table. The second option is to leave it in a regular table. The main issue with the flex table is that you can't tell the table the low values are good. This means that in order to have the…

Interesting. I wonder if this is a limitation of the "Other" bar. Could you try changing the beastmode to: (SUM(DATEDIFF(`Date Col1`,`Date Col2`)) / COUNT(`Date Col1`)) Are either of the two date columns ever null? You could also try: AVG(case when `Date Col1` is not NULL AND `Date Col2` is not NULL then DATEDIFF(`Date…

That's odd. Is the card built off the raw data? Or is there a dataflow that the card uses?