search

kaijagahm / ygdpDashboard

Interactive dashboard for YGDP survey data
3 stars 0 forks source link

(INT) recompute hexes to properly cover Michigan #22

Closed kaijagahm closed 3 years ago

kaijagahm commented 3 years ago

The way I computed the hexes, based on the boundaries of the continental United States, somehow missed a large portion of Michigan (at least for the large size hexes).

I haven't dug into why this happened. Presumably this could be fixed by adjusting some parameters in the idw_hex_interpolation.R script having to do with the smoothing of the minimum concave polygon (like, the concavity parameter? Did I even end up using the concavity thing for the final version of these analyses? I don't remember).

Including that one portion of Michigan also isn't the most critical thing in the world. But I think people will notice, so it would make sense to fix it at some point.

kaijagahm commented 3 years ago

Now using the medium size hexes, and the problem seems quite a bit better. I think this happened initially because I restricted the hex grid to only hexagons whose centers fell within the bounds of the continental US, and in the large size hexes, the center of the hexagon over MI didn't quite fall within the boundaries.

I'm going to close this issue, since using the medium grid fixes the problem. Could re-open an issue if it turns out that for some reason we have to use the large hexes.