3 ms·

Looks great, and generally very smooth for me. When the tooltip changes location it is quite jarring, but I have no good solution for that :) - perhaps only fl

by Cogito 13y ago

Looks great, and generally very smooth for me. When the tooltip changes location it is quite jarring, but I have no good solution for that :) - perhaps only flip the position when you would otherwise obscure the underlying visualisation, rather than when you cross the centre? You could then have the tooltip 'push up' to the edge when you navigate close to it, but when the tooltip would otherwise not be obscuring. There was a submission to HN not too long ago that featured a library that would simplify creating this kind of behaviour, but alas I cannot find it. It took me a while to click on the different years at the bottom, and until I did I wasn't certain what the tooltip data was representing. Probably not a problem, as the discovery was fun, but I guess many might not explore that far. Out of interest, what were your design thought process when choosing a map projection?

1wheel 13y ago
Didn't notice the tool jumping around till you pointed it out; now I can't stop seeing it. Pushing against left and bottom would definitely be better, especially since there isn't much to see there. Embarrassingly, I'm not sure what the projection is. The data is described as 'estimates of human population for the years 1990, 1995, and 2000 by 30 arc-second (1km) grid cells'. I did some aggregation to reduce the file size and then just converted that grid to pixel values stretched across the screen: y = d3.scale.linear() .domain([0, longitudes.length]) .range([0, height]), http://sedac.ciesin.columbia.edu/data/set/grump-v1-population-count http://sedac.ciesin.columbia.edu/data/set/grump-v1-populatio...
- Cogito 13y ago
  I found the projection on the GRUMP FAQ page[0]. Turns out they don't use a projection: *Projection: None, the data is in Geographic (latitude/longitude)* and so the resulting map is similar to the Mercator. I wonder how easy it is to map lat/long co-ordinates into different projections - would be a cool option to have on all similar maps. [0] http://sedac.uservoice.com/knowledgebase/articles/41635-what-is-the-projection-used-for-the-data http://sedac.uservoice.com/knowledgebase/articles/41635-what...
  - danpat 13y ago
    Use Proj.4: http://trac.osgeo.org/proj/wiki/man_proj http://trac.osgeo.org/proj/wiki/man_proj or, just take a raster image (PNG) and warp it: http://www.gdal.org/gdalwarp.html http://www.gdal.org/gdalwarp.html

by Cogito 13y ago

1wheel 13y ago

Didn't notice the tool jumping around till you pointed it out; now I can't stop seeing it. Pushing against left and bottom would definitely be better, especially since there isn't much to see there. Embarrassingly, I'm not sure what the projection is. The data is described as 'estimates of human population for the years 1990, 1995, and 2000 by 30 arc-second (1km) grid cells'. I did some aggregation to reduce the file size and then just converted that grid to pixel values stretched across the screen: y = d3.scale.linear() .domain([0, longitudes.length]) .range([0, height]), http://sedac.ciesin.columbia.edu/data/set/grump-v1-population-count http://sedac.ciesin.columbia.edu/data/set/grump-v1-populatio...

Cogito 13y ago
I found the projection on the GRUMP FAQ page[0]. Turns out they don't use a projection: *Projection: None, the data is in Geographic (latitude/longitude)* and so the resulting map is similar to the Mercator. I wonder how easy it is to map lat/long co-ordinates into different projections - would be a cool option to have on all similar maps. [0] http://sedac.uservoice.com/knowledgebase/articles/41635-what-is-the-projection-used-for-the-data http://sedac.uservoice.com/knowledgebase/articles/41635-what...
- danpat 13y ago
  Use Proj.4: http://trac.osgeo.org/proj/wiki/man_proj http://trac.osgeo.org/proj/wiki/man_proj or, just take a raster image (PNG) and warp it: http://www.gdal.org/gdalwarp.html http://www.gdal.org/gdalwarp.html