Skip to main content

Power BI

Needs Votes

Ability to use Trend Line with Categorical Values in X-axis

Vote (270) Share
's profile image

on 26 Sep 2016 21:17:59

Right now we can only use trend lines for continuous values in the x axis. It would be great if a categorical value could also be used, such as month name.

Comments (37)
's profile image Profile Picture

b06fca53 a9d2-4e9e-ae94-44ad0a4f6349 on 05 Jul 2020 23:50:05

RE: Ability to use Trend Line with Categorical Values in X-axis

This will help our organization so much!

's profile image Profile Picture

a0e22dda 0f15-4137-ad1c-db7b70a5c20d on 05 Jul 2020 23:50:01

RE: Ability to use Trend Line with Categorical Values in X-axis

This would greatly enhance the reporting capabilities at our agency!

's profile image Profile Picture

330e29ac ed43-4200-a796-bba2e5528049 on 05 Jul 2020 23:49:37

RE: Ability to use Trend Line with Categorical Values in X-axis

Linear trendlines need to be available in Power BI. Our healthcare organization is in the process of converting all of our control charts from QI Macros to Power BI. We were disappointed to learn that this functionality is not available in Power BI. We always apply trendlines to our charts to better display whether the pattern of data points is increasing or decreasing.

's profile image Profile Picture

43c1116e 4bfe-4fb8-b707-868bd622921e on 05 Jul 2020 23:49:35

RE: Ability to use Trend Line with Categorical Values in X-axis

This is extremely important for our organization

's profile image Profile Picture

164df75a d617-4b9d-9216-5b5158341e1d on 05 Jul 2020 23:48:00

RE: Ability to use Trend Line with Categorical Values in X-axis

I can understand why this doesn't work in its current state: in order to forecast/trend, Power BI needs to know relative distance between each category value. But regardless, a very common user story is not currently supported. Two possible solutions:
1) Use continuous (not categorical) but force Power BI to display all values of x-axis (not matter how cluttered that makes the visual)
OR
2) Activate Trendline for Categorical, but caveat to user that Power BI will infer equal increment across category values.

's profile image Profile Picture

9e4c10f0 daae-4bdd-98a9-6a1231dc7e18 on 05 Jul 2020 23:45:29

RE: Ability to use Trend Line with Categorical Values in X-axis

how is this not a thing?

's profile image Profile Picture

d32d8a66 7aa6-4e06-aacc-c88177adce3d on 05 Jul 2020 23:45:01

RE: Ability to use Trend Line with Categorical Values in X-axis

so much needed...

's profile image Profile Picture

ffc95a18 077c-4f05-875f-9b46c01a75f0 on 05 Jul 2020 23:43:47

RE: Ability to use Trend Line with Categorical Values in X-axis

this is an important feature which would make trend lines SO much easier. Currently not having a continuous option when plotting categories (even if those categories are a custom calendar e.g "2018 P1") is a big pitfall and makes trend lines impossible.

's profile image Profile Picture

4c9ff730 2d9c-4e5b-9886-abda7f8d03f8 on 05 Jul 2020 23:40:53

RE: Ability to use Trend Line with Categorical Values in X-axis

Agree - you have NO idea what cosmic workaround calculations me and my colleague made to get this thing working...

's profile image Profile Picture

5d8846b4 680e-4f5b-b207-ad682a443d79 on 05 Jul 2020 23:38:16

RE: Ability to use Trend Line with Categorical Values in X-axis

I don't even understand why this doesn't work this way already. It makes sense that the value the trend line is mapping would need to be numerical, but what does it matter how it's split up? If I have 5 number values in a chart, you shouldn't need to know what I call those groups in order to calculate a trend line... it's already divided into 5 discrete datapoints which you can use to generate a trend line. If your calculation needs numbers to map against, just use the column number!

For example, if I have a chart that that shows Page Views for:
"March"(55)
"April"(65)
"Cucumber"(42)
"Pickle"(99)

You don't need to calculate the value of "Pickle" to map a trend line, just read those values in as (1,55) (2,65) (3,42) (4,99) and calculate the trend line!