Great video but i am still not sure is when to use either one. specifically what the queues we need to pick up to decide between blend and joint. or we just have to test both to see which will provide the most accurate VIz and results. a bit tedious and time-consuming with the validation. maybe go with blend all the time when working with two data sets or databases. what do you think
So typically, joins solve most of life's problems haha and the way I think about it is you won't need a blend unless the joins aren't able to give you what you need. Typically, when you're joining to data sets together at different granularities i.e. one has sales by year, and the other has sales by day, this is when blends come in very handy. With joins, duplicates will occur whereas in blends they do not. Maybe I'll do another video demonstrating what I mean because blends are very tricky to see in your head!
So Blending and Relationships are essentially the same thing in my opinion. To be clear, we're not talking about Joins which I consider physical joins that result in a consolidated data set. So for a moment, forget what they're called. Instead, think of Blends and Relationships as doing the same thing two different ways. When you blend at the data source level, you explicitly tell Tableau how the data sets are connected. That way when you get to the visualisation stage its one less thing to think about when building your visualizations. On the other hand, when you set up relationships between data sources at the visualization level, you have tell it at each visualization you build, what the relationships are how the data is to be related. That's it in a nutshell. I think I'll create a video on this next week to really demonstrate it :)
Do you have anything on blends for older versions of Tableau? I'm a new Business Analyst and new to Tableau. I'm working with very large data sets that have already been curated by my IT department. My job is to add functionally to existing reports. My reports on Tableau server were created prior to relationships in Tableau. I'm having to work with a lot of data blends at the worksheet level. Still wrapping my head around blends, order of operations and level of detail.
Hey Phillip! It's probably best for me to have a look at the specific issue you have. All Blends are not create equally as they're used for different things. Would you mind getting in touch with me by email and I can take a closer look. guinto.jed@gmail.com
Great video but i am still not sure is when to use either one. specifically what the queues we need to pick up to decide between blend and joint. or we just have to test both to see which will provide the most accurate VIz and results. a bit tedious and time-consuming with the validation. maybe go with blend all the time when working with two data sets or databases. what do you think
So typically, joins solve most of life's problems haha and the way I think about it is you won't need a blend unless the joins aren't able to give you what you need.
Typically, when you're joining to data sets together at different granularities i.e. one has sales by year, and the other has sales by day, this is when blends come in very handy. With joins, duplicates will occur whereas in blends they do not. Maybe I'll do another video demonstrating what I mean because blends are very tricky to see in your head!
Hi Jellyman, But you are using relationship that is not a blending rite?
So Blending and Relationships are essentially the same thing in my opinion. To be clear, we're not talking about Joins which I consider physical joins that result in a consolidated data set.
So for a moment, forget what they're called. Instead, think of Blends and Relationships as doing the same thing two different ways.
When you blend at the data source level, you explicitly tell Tableau how the data sets are connected. That way when you get to the visualisation stage its one less thing to think about when building your visualizations.
On the other hand, when you set up relationships between data sources at the visualization level, you have tell it at each visualization you build, what the relationships are how the data is to be related.
That's it in a nutshell. I think I'll create a video on this next week to really demonstrate it :)
Do you have anything on blends for older versions of Tableau? I'm a new Business Analyst and new to Tableau. I'm working with very large data sets that have already been curated by my IT department. My job is to add functionally to existing reports. My reports on Tableau server were created prior to relationships in Tableau. I'm having to work with a lot of data blends at the worksheet level. Still wrapping my head around blends, order of operations and level of detail.
Hey Phillip! It's probably best for me to have a look at the specific issue you have. All Blends are not create equally as they're used for different things. Would you mind getting in touch with me by email and I can take a closer look. guinto.jed@gmail.com