DavidChurchman Coach

Have you identified that there are matching rows for each side of the join? Sometimes when troubleshooting a frustrating ETL, it's good to start with a very small subset of your full data. Here, you could filter both sides of your join for a value (or handful of values) that you believe should match, create an output for…

In that snip, there are values in the two date filters, which I believe is what is triggering the filter icons. Have you tried clearing those two filters? I don't think you can add any conditions to the filter icons, but you can turn them off for the whole page

For some reason, the styles applied in App Studio don't apply to the card when it is viewed in its expanded/detailed version (hopefully something they correct in the future). You can edit the alternating row colors within the card itself, or just turn them off in the General property settings:

This is not what you're asking for, but for a quick-and-dirty solution, I've starting using the Zoom Workplaces "Clips" to create quick video walkthroughs of my dashboards, and then I upload them directly to the dashboard as a document card so users don't have to go looking for them. (In App Studio, I put it on the last…

It's a little unclear on what you're trying to do, but since your step 1 and 2 are mutually exclusive based on their filters, I think you probably don't want a join, since there won't be any matches to join on. You probably want to combine the data using an "Append" tile instead, which stacks the data on top of itself.

You could round and multiply by 100 in a beastmode. (Not sure how many digits after 0 is useful for you, but probably not more than 2) ROUND(BM DIFF * 100, 2) Then you could edit your notification message to put a % after the values in the email.

I can confirm that I see that too. With no tiered dates, I'm able to select a range of columns: With the tiered dates, I am no longer able to select an area of the chart to filter (a picture showing an absence of a behavior, lol): For how helpful trellises can be in organizing a big chunk of data, there are some…

For #4, I would recommend the 'Split Column' tile, because it lets you keep the "discarded" characters in a separate column, in case you need to reference them later. (You can always drop the extra column using a the "Alter Columns" tile.

I don't think you can concatenate strings across rows using BeastModes, which seems to be the fundamental challenge here. I'm mostly commenting in case I'm wrong and someone posts a way to do that. Depending how many different types of filters you have and how many different product categories, you might be able to create…

I think you could do all of this with no BeastModes or Variables. A variable would be useful if you were trying to do this all with one card, with the ability to switch the metric from MTD, QTD, YTD, but you want to use three different cards, so you can easily define the "graph by" on each card using the date options. I…

%M is for full-month name "June". Try %b for three-letter abbreviated month name. That's as far as I got in checking, but I would double check the rest of your codes, too:

Make your font size bigger and your card shorter. The data table is pretty much fixed in size, but your bars will adjust based on the height of the card.

Not to gatekeep the fun that can be had in Jupyter, as someone that's describing themself as zero-code, I think ETLs and exploring the data with tables is probably a better starting point than trying to learn Python or R to do clustering. Since you said your dataset has hundreds of columns, I would recommend starting by…

I think it's because you have the "value format" set to "currency" instead of "percentage". This is from the documentation:

Is there a reason to do a distinct sum rather than a straight sum? There are probably non-distinct values that make up your monthly totals. Let's say July was: 100, 100, 100 for a total of 300; August was 50, 50, 50 for a total of 150, so your totals table would look like: July: 300 August: 150 Your overall distinct sum…

I don't think you can do exactly what you want, but it seems like you're trying to make a contour plot. Probably the closest thing Domo does to this that's relatively easy to set up is a Heat Map: If a scatterplot is important, you could also play with color coding your points to capture this effect as well:

Make sure you check this box in the "general" properties.

Yeah, the null handling ain't great. If you're using an ETL and have access to the R/Python tiles, you could handle all your nulls in all your columns with a quick script: library(dplyr) library(tidyr) library('domomagic') input1 <- read.dataframe('Alter Columns 2') input1 <- input1 %>% mutate( across(everything(),…

Yes, I love this. Even doing something like switching between different types of bar charts based on variables would be great. Some variables might be better as 100% bars, others as grouped and others as stacked.

For your second question, I believe you could just wrap your case statement in the other David C's formula: SUM(COUNT(DISTINCT [YOUR CASE STAMENENT HERE])) over (partition by group) But for your first question, I'm not sure I understand what you want for the totals of your % column. You have OwnerIDs that fall across…

You could do this with a Histogram (though the labels would be slightly different than your picture). "Category Names" would be the phone numbers and "Column containing data set" could be a beastmode like COUNT(TransactionDate) or COUNT(DISTINCT TransactionDate) To keep it as a horizontal bar chart, I would use two…

Maybe you can use a fake dataset and post what you're seeing and talk through why it's not pretty? I've been playing with this example as a test case. I tend to find the tables in Domo to be ugly/difficult to make functional. The new border options help a fair amount, but it still feels like I have to hack it a bit to look…

I don't know a way to do this within a single card, but it would be relatively easy to have an aggregation card next to your table card that has both of these values by category.

In your opening post, there is no sum() function around quantity. You tried it with that sum function?

I can't recreate this. Changing the custom colors in App Studio does change the colors in a line chart. It looks to me like you have "Gradient" colors turned on for the card, so I would double check that, because any color settings manually entered in the card will overwrite any styles applied by the color palette in App…

Pivot cards don't play well with HTML formatting, but they can accept unicode or emojis, which can serve as indicators that you can manipulate in a case statement. For example, I made this pivot card: Using this case statement: case when Category = 1 and sum(Orders)/sum(sum(Orders) fixed(by week No.)) <0.05 then…

Colemen's answer will give an empty string instead of a NULL. If you want a NULL, then you could flip your CASE statement. Everything that is not "not 1" will be NULL: CASE WHEN Sort: Month Number != 1 THEN ( (SUM(All Cases) - LAG(SUM(All Cases), 1) OVER (ORDER BY Sort: Month Number)) / LAG(SUM(All Cases), 1) OVER (ORDER…

Hey, we all need a juice-box and a nap sometimes

When you said "some event to occur" in the OP, I didn't know that's what you wanted. If you've shared that summary number to the dataset, for any cards that draw from the same dataset, you could use card-level filters that require that field to be above 10, and then you can specify a "No data" message.

I would look at your History to see if it's truly running when it says its running. Also, double check your Versions. What was the difference between the last version and this one? Do you have anything unconnected in the ETL, maybe hiding in some corner?