Find everything you need to get certified on Fabric—skills challenges, live sessions, exam prep, role guidance, and a 50 percent discount on exams.
Get startedEarn a 50% discount on the DP-600 certification exam by completing the Fabric 30 Days to Learn It challenge.
In the R and Python languages there exist packages such as caret/ggplot2 [ R ] and seaborn [ Python ] for creating scatter plot matrixes that show you a bunch of dataset feature variables, e.g. the variables that could contribute to predicting a single variable of interest, on individual scatter plots against each the other feature varialbes and the label variable, i.e. the variable you are looking to predict using feature variables.
The objective is to do an exploratory data analysis visually to see which feature variables have a good linear relationship with the others and specifically the label variable. In the example below the plots in the matrix diagnal are histograms of each of the feature variables and the label variable which helps with doing a quick initial visual review of distributions and cases with outliers.
Does power bi have a way, hopefuly easy, to create this kind of scatter plot matrix as is shown in image below?
You can create R visuals inside Power BI, and you can probably use exactly the same code in Power BI that you used to create that screenshot
Have a read of this
https://docs.microsoft.com/en-us/power-bi/desktop-r-visuals
Thanks for response and suggestion. Yes i am aware of the that report control that allows you to incorporate R script directly to generate a report view. That said what i'd like to avoid is having to resort to R [ or Python ] script for generating visuals whereever possible and leverage power BI's feature set for accomplishing similar things.
I've tested the simple correlation matrix visual outlined as the example in the article you provided link to, i.e.
require(corrplot)
M <- cor(dataset[c('geo_level_1_id', 'geo_level_2_id', 'geo_level_3_id', 'age', 'area', 'height', 'damage_grade')])
corrplot(M, method = 'color', tl.cex = 0.6, tl.srt = 45, tl.col = 'black')
and that is great for case where R generated visual generating method call is easy to use and configure the output to look nice.
On similar front, see http://community.powerbi.com/t5/Desktop/power-bi-for-pretty-histograms/m-p/391334, its a shame that to create something as common as a histogram / frequency chart in power BI i have to go off into the query editor and create a dataset with group by column and associated count of unique values for every column[/field/variable] that i want to produce a histogram for in my report view. I would have expected that common exploratory data analysis visual to be a dragNdrop exercise unless it is and i've overlooked how to do it. Resorting to R script control to accomplish that means i'm now relegated to R hist() services for making it pretty which isn't as quickNeasy and maintainable by non-R users as doing it in power Bi would be.