Strange behavior on ggplot2

Question

I'm trying to do a map to identify specific areas by coloring them. First, I made this plot to check if the data was ok (Setor is the sector's number):

ggplot(aes(x = long, y = lat, fill = Setor), data = mapa2010) + geom_polygon(colour = 'black') # data is ok

Them I tried to made the plot, filling by another variable (AGSN):

ggplot(aes(x = long, y = lat, fill = AGSN), data = mapa2010) + geom_polygon(colour = 'black')

The data is exactly the same, there is no code lines between this 2 commands. I've already tried to reorder the data, but still wrong.

Anyone know why this happens, and how to solve it?

[How to make a great R reproducible example?](http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) — zx8754, Oct 08 '15 at 16:49
It's hard to do reproducible example because I don't know why this happens (to recreate in the example), and the original data is big — Rcoster, Oct 08 '15 at 16:52
Looks like rows in your data frame got re-ordered (mis-ordered). — Gregor Thomas, Oct 08 '15 at 17:13
@Gregor: I agree with you. But why this happens? I tried sort the data by: `AGSN`, `id`, `piece` and `order`, but didn't changed anything. — Rcoster, Oct 08 '15 at 17:59
Do you have missing values in `AGSN`? If so you probably got a warning with your plot, but that could do it... — Gregor Thomas, Oct 08 '15 at 18:03
No missing values. Don't know why, but the second map requires `group = group` in `aes()`. — Rcoster, Oct 08 '15 at 18:14

score 1 · Accepted Answer · answered Oct 08 '15 at 18:16

Adding the parameter group = group in aes() for second plot solve. Don't know why only the second map needs.

ggplot(aes(x = long, y = lat, fill = AGSN, group = group), data = mapa2010[order(AGSN, id, piece, order), ]) + geom_polygon(colour = 'black')

Strange behavior on ggplot2

1 Answers1