Venn Diagram

Hi Anonymous, 

I am glad it worked out for you. Here are limitations of R visuals in PBI: 
https://docs.microsoft.com/en-us/power-bi/desktop-r-visuals#known-limitations

HI Again Boefraty,

I did everything you asked, but now I'm getting a count of one per category. There is a unique id for each observation, but Power Bi is still deleting "duplicates".

Here's the code and output, as well as a screenshot showing the unique ids.

Many, many, many thanks,

Juniper

#set working directory i.e. the folder you are working out of
setwd("H:/Eliel JBBS Sent")
#set default library path - where to look for packages, which are also called libraries
.libPaths("H:/R/win-library/3.4")

#load Venn Diagram package

library(VennDiagram)
library("RColorBrewer")
library(grid)
library(futile.logger)

#Venn <- read.xlsx("Venn.xlsx", sheetName = "Venn", header = TRUE, stringsAsFactors = FALSE)
Venn = dataset

draw.triple.venn(margin = 0.1, area1 = nrow(subset(Venn, Felony == "Yes")), area2 = nrow(subset(Venn, Misdemeanor == "Yes")), area3 = nrow(subset(Venn, Ordinance == "Yes")),
n12 = nrow(subset(Venn, Felony == "Yes" & Misdemeanor == "Yes")), n23 = nrow(subset(Venn, Misdemeanor == "Yes" & Ordinance == "Yes")),
n13 = nrow(subset(Venn, Felony == "Yes" & Ordinance == "Yes")), n123 = nrow(subset(Venn, Felony == "Yes" & Misdemeanor == "Yes" & Ordinance == "Yes")),
category = c("Felony Offenders", "Misdemeanor Offenders", "Ordinance Offenders"), lty = "blank", fill = brewer.pal(n = 3, name = "Dark2"))

Hi Anonymous, 

I  really need the PBIX to see  the problem  :(

Sounds like, you added the unique ID to data table, but you did not add the ID column as input to R visual. Drag the ID column to R visual as you did with "Felony" and the rest.  

Hi Boefraty,

Thank you so much for your patience. Here is the PBIX file.

I dragged the unique ID column to the R visual, but got the same result.

Juniper

You missed my "tip" about making sure each column to be  "Don't summarize" . Working PBIX with filters is also attached 

Hi Boefraty,

That's so strange. My file on this end says "don't summarize" is selected. Your code won't run because it doesn't like the (perfectly good) file path, and it says duplicate rows have been removed from the data, which isn't supposed to happen.

Many thanks,

Jennifer

Dear Boefraty,

I got it to work! Thanks so much. You have been amazing. I truly appreciate your help and your patience.

Thanks again,

Juniper

Hi Boefraty,

How large of a data set can a Venn Diagram handle in Power Bi?

Many thanks,

Juniper

Thank you so much!!!

This visual appears to give incorrect results when 4 fields are selected.

I can do 3 just fine but 4 doesn't calculate properly, e.g., a category with 30 items shows 20, 4, 0, 0 instead of adding to 30.

My math could be off, though. I think with 4 categories there's no way to actually show the top & bottom categories (or left & right) intersecting since they already overlap in the middle.  Maybe that's just a limitation of Venn diagrams; in that case maybe limit it to 3 categories?

3 categories: correct figures; each category adds to 8.4 categories: Something is wrong; categories add to 7.

Data is standard 4 variable w/ 2 options per variable.

Hi spotpuff, 

Please make sure all the rows in your input table are UNIQUE. You can do it by adding ID column (but ignore it in R script). 

The rows are unique; I was using an illustrative example but I was using the diagram on a real dataset which has a row ID type field and I was still having problems with the 4 dimensional Venn diagram.

The issue is likely something to do with more than 3 categories requiring strange diagrams:

https://en.wikipedia.org/wiki/Venn_diagram#Extensions_to_higher_numbers_of_sets

The visual doesn't make the 4th category into that "rainbow" type thing, so the math might be off.

Or I might be wrong.

Either way for now I'm sticking with 3 categories as the math works out in that diagram.

Hi spotpuff

The code uses the VennDiagram package, check their documentation. 

Can I get the PBIX with exact R code and data which you use for testing ?