Suppose we want to reflect on a map, the electoral results of the municipalities, so that they can be filtered by political party and shared with the public. While there are more less profane ways to do it, I want to show the example to explain how it can be done with FusionTables by an ordinary user.
What we have:
The published result of the Supreme Electoral Tribunal, where you can see the listing by municipality.
Minute 1. Build the table
This is done by copying and pasting from the table available by the Supreme Electoral Tribunal to Excel. Special copying is used, only text and as there is no country display, you have to filter for each of the 18 departments. The advantage with Chrome is that the selection is made, even if we change the filter so that we only have to do Ctrl + C.
We leave the header only in the first row.
As the table does not have a coordinate, it will be necessary to georeference it using geocode. To do this, we will concatenate the columns so that Google does not get confused when searching for the locations; we require you to look for a municipality, department, country.
In column F, we will use the concatenate formula: = CONCATENATE (Township column, ",",Department Column, ",","Country«), We are also concatenating commas between quotation marks to ensure that the chain is as we expect it. Thus, the column of row 2 would look like this:
= CONCATENATE (B2, »,», A2, »,», »honduras») and as a result that row would be: Central District, Francisco Morazán, Honduras
To the heading of this column E we will call it "Concatenate"
Minute 5. How to upload it to FusionTables
FusionTables is installed in the Google Chrome browser, and when you call it to create a new sheet from this link, This panel should appear.
You can choose a sheet available in Google Spreadsheets, create a blank or upload the one we have on the computer.
Once selected, select the "Next" button. It will ask us if in the first row the name of the columns is found, then we do «Next» and then it will ask us what name we give to the table and some descriptions that can be edited afterwards.
Minute 7. How to georeference the table
From the File tab, select the "Geocode ..." option and ask us which column contains the geocode. We indicate the column that we have defined before.
If we had not created a concatenated column, we could have defined the municipality but because there were so many repeated names in many countries, scattered points would have come out of Honduras. Also within the same country there are municipalities with the same name, for example "San Marcos", if we did not concatenate the department we would also have had that difficulty.
There is an option called "ad location hint", which in this case is not necessary because the entire chain already contains the information up to country level.
The system starts locating each location based on the criteria we have defined. Below indicates in orange the percentage of ambiguous data, which will usually happen with locations that Google has not yet identified in its database; in my 298 case, only ambiguous 6 came out; Usually Google places them in another country because they exist somewhere.
Minute 10, there they have it
If a point went out of place, it is edited in the "Row" option, by double clicking on the field and on the "edit geocode" link, improving the search and indicating the place that resolves the ambiguity. If it does not exist, then you can indicate a nearby location that we see on Google tags.
In the option of filters, it is possible to add panels to turn on, turn off and count by match, by department, by municipality ... etc.
Here you can see the example. It does not have the final data because I did it with information that was still being processed, also moving some tables making a merger with area and municipality code from another table ... but as an example there is the link. I did not do the internal correction for an elementary error and waiting for the 10 minutes to be enough.
You can merge tables, edit directly, publish and some other basic things. To do more, there is the API.
Of course, this is done by means of points.
If you want to use shapes for the columns you can use the Shapescape service (hopefully it is not down) ... although you want more than 10 minutes.