Check for missing data on Map Visual

[Anonymous]

Hi all,

I have a dataset with an address column. I'm plotting this information on the standard Map visual in Power BI. 

the address is formatted like this ([street number], [postal code] [city]) , eg: Lansstraße 81, D-11179 Berlin

Now, by accident, I've noticed that (at least) one address isn't plotted on the map (because the Bing map can't find the address maybe??). This address is certainly a correct address.

Is there any way I could determine if (and wich data) Power BI is refusing to plot on the map visual? I Can't check them one by one because there are over 1000 addresses in my dataset.

Thanks!

[v-xuding-msft · Accepted solution]

Hi @Anonymous ,

Please change the format ([street number], [postal code] [city]) to ([street number],[postal code],[city] and change the data catogory to Place to have a try.

If this can't work, please split the column from one column ([street number], [postal code] [city]) to three columns  ([street number]) ([postal code]) ([city]) and change their data catories to Place, Postal code and City.  Then drag all of them into Location field.

https://docs.microsoft.com/en-us/power-bi/visuals/power-bi-map-tips-and-tricks#in-the-dataset-tips-to-improve-the-underlying-dataset

[lbendlin]

You may also want to strip the "D-"  from the postal code - it's not officially part of it but rather identifies the Country (Deutschland). 

It may also help to use the right address.  According to Google the ZIP code for Lansstrasse 81 is 14195, but Street View reveals that Lansstrasse actually ends at house number 15.  So maybe number 81 is fictitious.

Yeah, looks like a fake:  https://microjobengine.enginethemes.com/blog/mjob_post/i-will-act-as-a-user-and-test-your-website/

[lbendlin]

Short of reverse engineering the map visual code you have not a lot of other options.  One option would be in Power Query (or any other tool) to add a Bing address resolution API call against each address and record the result.  

You wouldn't want to run that check all the time though - not sure how long it would take to check them all individually.  Maybe keep a master table with addresses and their last validation date, and then revalidate only the once that haven't been checked for more than a month.