Product Tour
forums
Use these forums to ask questions and discuss Tableau.
Light GIS (Geographic Information System)
You show examples with kludgy maps and simulated mapping data. I've had years of experience with MapInfo which is BI specifically for geographic data.
No other BI tool I've seen wants to deal with geography and this would be a huge differentiator. I believe the new SQL2008 explicitly ads support for many geographic aspects. It seems you've been teasing with it but not gone full guns.
It would be great if you included a base set of maps, down to say country, state county level and with zip code points. Then you could map your data and zoom in and out as well. One cool feature of Mapinfo was thematic maps where the state's (or any geographic boundary) color would change based on the underlying related data. You could then resell map data for more detail (such as streets) via an oem deal with a major provider.
Another nice visualization was pie charts. (Which you could also use on non map analysis). Basically the position of the pie chart on the map represents the location, the size of the pie can represent the total value, and the various wedges represents the values of the pieces that make up the total. This shows alot of information in a very intuitive way.
Perhaps you could integrate google maps into this in some way.
Comments
Have you been peaking through our office windows looking at our conference room whiteboards?
Stay tuned, we have good stuff coming on this front...
On this same theme, a challenge we have as a customer is when we map data, say for instance where patients live relative to the clinic they visit, it is beyond our abilities to somehow apply logic that identifies those whose point of origin are say more than 5 miles from the clinic.
Recalling my math classes, I can tell the difference between the patient's home latitude and the clinic latitude, I can tell the difference between the patient's home longitude and the clinic longitude, so in theory the length of the hypotenuse would be the distance travelled. I just don't know how to convert lat/longitude to miles to get a result.
Any "built-in" means for computing this would mean we as customers can skip the need to get a doctorate in mathematics or GIS. That'd be a nice to have for sure.
Hi Dan,
The following will calculate the distance between two points. First, convert the lats/lons to decimal degrees, which, if you're "mapping" in Tableau is already done. Next, convert the lats/lons into radians by multiplying them by pi/180 or 0.017453293. Now you've got the two pairs in radians and can perform a Haversine distance calculation as follows:
dlon = lon2 - lon1
dlat = lat2 - lat1
a = sin^2(dlat/2) + cos(lat1) * cos(lat2) * sin^2(dlon/2)
c = 2 * arcsin(min(1,sqrt(a)))
distance = R * c
If you want your distance in miles use R=3956 if you want distance in kilometers use R=6367. Hope this helps.
BTW, don't know much about MapInfo integration options but both the SAS BI tools and Spotfire integrate with ArcGIS Server. At InfoVis I mentioned to Jock that a Tableau - ArcGIS integration would be very useful.