Updating dataflow without updating datasource

Reply
Yellow Belt

Updating dataflow without updating datasource

Hi there!

 

I have troubles with DOMO and can´t figure it out without adding (hopefully) unnecessary data.

 

We run a daily update on costs on an external service via Excel upload.

In dataflow I join daily costs from the previous year.

We don´t have costs on every day, but we still need to run it, so costs from last year are included.

I use a schedule in dataflow, so data will be still updated without updating the datasource.

 

to clarify:

 

 DOMO - Updated Version DATAFLOW.jpg

As you can see we don´t have costs for December 18, 2018, but we do have costs for December 19, 2017, so this row must be included in the output.

Within dataflow I can see the correct figures, but in the output I get the following result:

DOMO output.jpg

 

On the 6th of december I updated the datasource the last time. But I set the update trigger to schedule so the data shouldn´t stop at the 6th.

 

Could someone explain me if I need to change some settings or something in DOMO?

 

Many thanks

 

 

Green Belt

Re: Updating dataflow without updating datasource

Hi, @Nikse,

 

I'm happy to help, but I don't think I fully understand your question. Let me summarize what I understand (also one question), and then please correct me.

 

  • You have a sparse cost per day dataset, meaning a record only exists for a particular date if a cost was incurred on that date
  • Since costs are not incurred every day, the dataset may not update every day
  • You want to compare cost per day of the current year with the equivalent day of last year
  • If there was no cost on a day this year but there was a cost on the equivalent day last year, the output should include a record for the day this year with a cost value of 0 with the equivalent day last year with its cost value
  • And vice versa in the case where this is a cost on a day this year but not last year
  • Seems that you have accomplished all this expected behavior
  • Do you have some logic in place in the dataflow to only output records where date of the current year is less than the current date?
    • If yes, you might consider applying this filter in the visualization layer (in the card; perhaps using a Beast Mode calculation) rather than in the dataflow. This way, you can set your process to run only when the source dataset is updated. That's because, when you run the dataflow on Dec 6, 2018, it will include records into the future for dates that have costs on equivalent dates last year.

 

I wonder if the output dataset hasn't updated since 12/6 because there hasn't been an update to the source dataset since then. You should be able to tell what is happening with the dataflow by clicking into the dataflow details and then looking at the "History" tab. If you mouse over the "..." button, it should give you some detail about the result of the dataset run. Additionally, you can click on any records of a dataflow run to see yet more details.

 

I hope this is helpful - but I might have missed the point entirely! Smiley LOL

 

Best,

Dan

Announcements
Win free lodging at Domopalooza! We want to hear your Domo story, enter our Dojo contest and win up to four nights lodging at Domopalooza. Click here!