Zip2Fips - map card issues

Jbrorby

When I try using a United States map card or a State map card, I get the result in the screen shot I've attached. In this example, I've narrowed it down to 2 donors from Illinois (shown in the data table below the card).

I am using the column "zip2fips.donor_state_tz" column for the "select county fips code" for US state map card., and for "select state name" column in United States map card.

"zip2fips.donor_state_tz" is just the "STATE" column from the zip2fips set, just renamed.

I am joining zip2fips in ETL on zip_String_leading_zero to donor.zip. Because in our database, they will lead with zero when applicable. Also, some of our donor zips include the extra 4 digits, so I trim them to the first 5 before they enter Domo.

If I select the chart property to hide no value items, the entire map disappears, but the summary number remains, so its as if something with the state values just isn't linking up with the card.

Also, if I use the non zip2fips column "donor.state", the card works fine. But not at the county level, probably because there can be more than one county with the same name.

I use this exact same join in a separate dataset, and I don't get this issue when using that dataset. The only difference is, in the dataset it works, I do not rename the zip2fip columns. I don't know why that would cause any issues, but it is the only thing I can think of that is different between the working and non-working.

Any thoughts?

Thanks

Dojo map card.PNG

MarkSnodgrass

Have you tried dragging one of your zip code fields or the actual fips code into the "select county fips code" field? It is expecting one of those values rather than the state name to properly populate. I would also suggest select the aggregation type to SUM for the field in the "select county value" field.

Hope this helps.

Jbrorby

I went ahead and tried all the zip2fips columns in both United States map card and US State map card and still had no luck. Just to test, I went into the ETL and removed the column name change and reverted them to their natural name. Then they started working properly. So that is odd and I may need to submit a help desk ticket. This wouldn't be too big of a deal but we use these columns for more than one location piece.

MarkSnodgrass

That is interesting and does sound like a bug that is worth reporting. You could use the formula tile in the ETL to have one column that is the original name and one column that is named for how you are using it in other places.

Jbrorby

I just wanted to follow up on this incase anyone else has a related issue. This was not a bug. The problem was that I was prefacing the columns with "zip2fip." I followed the advice below it it resolved my issue.

"I will provide some clarification. The US map is unique in that it really could be one of several different maps. The actual map loaded is inferred from the column name / alias (Data Label) of the first column. We first look to see if the column name includes "zip", if so we try to use the zipcode level map of the US. If it doesn't include "zip" then we look for "fip" or "county" in the column name, and if either of those are present we load the county level map of the US. We then do the same tests for the alias or "Data Label" for the column. Finally if the column name includes "st" short for "state" then we load the state level map.

So in this case, because you have set the column name in the ETL to be "zip2fip"... it picks up on that column name and determines the map to use should be the zipcode level map.

Two easy fixes.

1. Have the column name in ETL not include zip or fip as part of the name, but include "st"

2. Create a beast mode that references the ETL column where the beastmode name does not have zip or fip in it, but does have "st". EG "State_bm" = zip2fip.donor_state_troubleshooting"