08-14-2016 13:16 PM - last edited 11-16-2016 01:27 AM
This is great! Appreciate the time and effort it took to build and share this out!
I've been exploring the forecasting using the R integration and PowerBI, and was wondering if there was a way to explicitly show the predicted values and confidence intervals in a dynamic way. For example, in this situation, if I clicked on "Cuba" for this specific forecast example, how would I obtain the values shown in the R visualization on the Dashboard?
Thanks for the great visual.
I am getting weird x Axis (date) labels. 2016.2, 2016.4 and they don't seem to line up with February 2016 and April 2016. Attached is a png image of what i am seeing. I also included a standard pbi visual to show the correct dates.
Am I missing a formatting option?
Attached is a copy of the actual data used in the original births.csv file which I obtained from the pbix.
I got this by simply creating a new table on the report page, added all columns with no summarisation then used the export data option from that table.
As dates are exported from Power BI using a date-time format I used excel to reformat the Calendar Month column as just a date. I deliberately used a US format (mm//dd/yyyy) which might be more widely usable though this should automatically be recognised by Power BI anyway. If it is not and you use something different then just format in excel to suit your needs.
Obvioulsy this doesn't really gain you anything since the information was all already embedded in the pbix but it does allow you to fix the errors seen when looking at queries.
Of course, the technique is something that might be useful in other situations where you are missing a source file and want to recreate it so hopefully this helps someone.
(attached as a zip file since csv attachments are not supported on the forum)
First of all, awesome visual! It took me a week to script the ets forecasting in R on my data, and I've been able to get what I want, but then the problem is that it generates a completely new set of data, unrealted to the existing model in Power BI. Your visual solves this problem by integrating the R script (very similar to what I have) into Power BI, so now I can use measures. Great work.
Afew things missing here: the Y-axis is poorly marked, if we're talking about millions of dollars on the Y axis. Also, no labels. So, I decided to clone this visual and create my small variation. Standing on the shoulders of giants, as they say
All I've done so far is adjusted the R script to also populate a table of values along with the graph. I left the rest untouched.
I got to the point that suits me for now (I hope to keep improving) and I can see needed results on the web when I run pbiviz start command.
But when I try to package the visual, the packaged custom visual for PBI Desktop doesn't have any of those changes... What am I doing wrong?