Best Of
Re: Totalling certain rows
@jakebutterfield If you want to show the totals for each year, add a beast mode using the formula YEAR(`Date`) to extract the year from the date column. Then add that new beast mode to the pivot table in your second image in the rows in front of Date. The pivot table will add subtotals by year once you do this.
Re: Card not reflecting updated Dataset/Dataflow updates
@MarkSnodgrass Thanks. Once removing and adding back the filters, the visualizations updated.
Re: Selecting only newer records from a remote datasource to append to a Domo datasource
@MichelleH Yup… I'm rolling 4 days worth of data and merging into DOMO. This should cover me for outage over the W/E maybe. As the data is an audit. Most of the historical data selected should be ignored and only new records inserted from the last inserted record.
Seems to be working well. Thanks
JohnnyN
Re: DataSet Views - Comments & History
@kensie_rowell I like the idea of tracking user and adding comments. But there is currently pretty extensive change long available (just hard to find and no user recorded or shown). If you click on the round arrow clock in top right of the edit screen you have details of what has changed. see attachments
Re: How to create a stacked bar chart?
@MichelleH I added a tooltip to the chart. In doing so, the option to add a series appeared, so I added the series and deleted the tooltip.
Re: Sum column when i have duplicates in other two columns?
So in an group-by tile, add Publication and InHomeYear to the groupings and do a SUM(Total Cost) in the bottom section which should result in what you are trying to achieve
Re: How to input identifiers using identifiers
Just to be clear, your dataset currently has 3 columns:
- row number (unique always filled)
- C which currently has only a couple of values and most of them are null, and that you're trying to fill out
- Property Data with the actual names of the companies
While I would strongly advise to have this filled in the source, if that is not an option, this can be done via MagicETL in multiple ways. Myself, I'd do:
Get only the rows that have a C value, append a JoinKey of 1 on both the filtered and the full set, do an Inner Join using the key, filter the results only to include those where the original row number is less than or equal to the joined row number, then do a Rank on the joined Row number in descending order, partitioning by the row number, filer the output to only include Rank = 1 and alter the columns, renaming the Joined C to C and dropping the Rank, JoinKey, Joined row number, Joined Property Data and C, finally store this to a new dataset. This will give you the output you're looking for, without having to use Lag an unknown number of times.
This will work regardless of how many blanks you have and will always fill the blanks with the values of the last known one, the only dangerous thing is if the lowest row number on your dataset is lacking a value for C then all rows until the first appearance of a value for C will be removed. Here's a side by side from the source and the output:
Re: Feedback Survey data from Hubspot into Domo
Just for reference in case anyone else is looking for this. As Grant said not available in normal Hubspot connector. I ended up using a workflow to automate a Webhook push from Hubspot to a Domo JSON Webhook connector and was able to get the data into domo that way
Re: Date format question
Agree with @MichelleH, first thing is to know what valid values look like in Domo, as there is no intrinsic conversion from Float to Date. Since the type is returned as float, that either means most of your entries were -3 which is how Domo decided to use that type for that field (not so great news and perhaps you'd need to research why -3 is being used in Snowflake for these) or that Snowflake returns the date as a number, perhaps as Epoch, in which case you'll need to convert it using FROM_UNIXTIME() and a base 10 factor.
Re: Group by when using case statements?
One solution could be to adjust the chart properties. For example, if I use the Maximum slices before 'Other' property from the data you provided I can change the graph to show the desired result:
The advantage to this, is that you can mouse over the "Other" pie and see what it contains as well:
