ggplot not plotting zeros and missing values

Question

I'm trying to make a map of Kenya with a fill that constitutes of values '0-3' and '5'. However, ggplot is plotting the counties with a '0' fill as blank and shifting the values on the scale, because there is no value '4'. How can I fix this? In the way that the zeros are plotted and the scale stays correct. Below is the code I used to make the plot. And also the code that I used to make the dataframe for the plot.

Let me know if you want me to add something and thanks in advance!

These are the first 10 rows of my dataframe:

structure(list(long = c(35.78634226, 35.79574167, 35.8581196, 
35.86048445, 35.88332712, 35.88631784, 35.90456735, 35.9095112, 
35.91591989, 35.93612253), lat = c(1.65376487, 1.63856712, 1.54176536, 
1.53814079, 1.50313011, 1.49812523, 1.46913353, 1.46174827, 1.4513723, 
1.41957298), order = 1:10, hole = c(FALSE, FALSE, FALSE, FALSE, 
FALSE, FALSE, FALSE, FALSE, FALSE, FALSE), piece = structure(c(1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = c("1", "2", "3", 
"4", "5", "6", "7", "8", "9", "10", "11", "12", "13", "14", "15", 
"16", "17", "18", "19", "20", "21", "22", "23", "24", "25", "26", 
"27", "28", "29", "30", "31", "32", "33", "34", "35", "36", "37", 
"38", "39", "40", "41", "42", "43", "44", "45", "46", "47", "48", 
"49", "50", "51", "52", "53", "54", "55", "56", "57", "58", "59", 
"60", "61", "62", "63", "64", "65", "66", "67", "68", "69", "70", 
"71", "72"), class = "factor"), id = c("Baringo", "Baringo", 
"Baringo", "Baringo", "Baringo", "Baringo", "Baringo", "Baringo", 
"Baringo", "Baringo"), group = structure(c(1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L), .Label = c("Baringo.1", "Bomet.1", "Bungoma.1", 
"Busia.1", "Elgeyo-Marakwet.1", "Embu.1", "Garissa.1", "Homa Bay.1", 
"Isiolo.1", "Kajiado.1", "Kakamega.1", "Kericho.1", "Kiambu.1", 
"Kilifi.1", "Kilifi.2", "Kilifi.3", "Kilifi.4", "Kilifi.5", "Kilifi.6", 
"Kilifi.7", "Kilifi.8", "Kilifi.9", "Kilifi.10", "Kilifi.11", 
"Kilifi.12", "Kilifi.13", "Kilifi.14", "Kilifi.15", "Kilifi.16", 
"Kilifi.17", "Kilifi.18", "Kilifi.19", "Kilifi.20", "Kilifi.21", 
"Kilifi.22", "Kilifi.23", "Kilifi.24", "Kilifi.25", "Kilifi.26", 
"Kilifi.27", "Kirinyaga.1", "Kisii.1", "Kisumu.1", "Kitui.1", 
"Kwale.1", "Kwale.2", "Kwale.3", "Kwale.4", "Laikipia.1", "Lamu.1", 
"Lamu.2", "Lamu.3", "Lamu.4", "Lamu.5", "Lamu.6", "Lamu.7", "Lamu.8", 
"Lamu.9", "Lamu.10", "Lamu.11", "Lamu.12", "Lamu.13", "Lamu.14", 
"Lamu.15", "Lamu.16", "Lamu.17", "Lamu.18", "Lamu.19", "Lamu.20", 
"Lamu.21", "Lamu.22", "Lamu.23", "Lamu.24", "Lamu.25", "Lamu.26", 
"Lamu.27", "Lamu.28", "Lamu.29", "Lamu.30", "Lamu.31", "Lamu.32", 
"Lamu.33", "Lamu.34", "Lamu.35", "Lamu.36", "Lamu.37", "Lamu.38", 
"Lamu.39", "Lamu.40", "Lamu.41", "Lamu.42", "Lamu.43", "Lamu.44", 
"Lamu.45", "Lamu.46", "Lamu.47", "Lamu.48", "Lamu.49", "Lamu.50", 
"Lamu.51", "Lamu.52", "Lamu.53", "Lamu.54", "Lamu.55", "Lamu.56", 
"Lamu.57", "Lamu.58", "Lamu.59", "Lamu.60", "Lamu.61", "Lamu.62", 
"Lamu.63", "Lamu.64", "Lamu.65", "Lamu.66", "Lamu.67", "Lamu.68", 
"Lamu.69", "Lamu.70", "Lamu.71", "Lamu.72", "Machakos.1", "Makueni.1", 
"Mandera.1", "Marsabit.1", "Meru.1", "Migori.1", "Mombasa.1", 
"Mombasa.2", "Mombasa.3", "Mombasa.4", "Mombasa.5", "Murang'a.1", 
"Nairobi.1", "Nakuru.1", "Nandi.1", "Narok.1", "Nyamira.1", "Nyandarua.1", 
"Nyeri.1", "Samburu.1", "Siaya.1", "Taita Taveta.1", "Tana River.1", 
"Tharaka Nithi.1", "Trans Nzoia.1", "Turkana.1", "Uasin Gishu.1", 
"Vihiga.1", "Wajir.1", "West Pokot.1"), class = "factor"), SUB_crisis_trans_count = c(1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L)), row.names = c(NA, 10L), class = "data.frame")

SUB_count_cristrans_KE <- long.SUB_dfCSKE %>% mutate(crisis = ifelse(`IPC class` %in% 3:5, 1, 0)) %>%
  arrange(`County`, `Period of measurement Kenya`) %>%
  group_by(`County`) %>%
  summarize(SUB_crisis_trans_count = sum(diff(crisis) > 0))

### Join the dataframe with the counts with the County polygon

SUB_cristrans_KE_plot <- SUB_KE_fortified %>% left_join(SUB_count_cristrans_KE, by = c("id" = "County"))

### Make labels for the map

minVal_cristrans_KE <- min(SUB_cristrans_KE_plot$SUB_crisis_trans_count, na.rm = T)
maxVal_cristrans_KE <- max(SUB_cristrans_KE_plot$SUB_crisis_trans_count, na.rm = T)
### Plot the crisis transitions per county of Kenya

SUB_Kenya_cristrans_plot <- ggplot() +
  geom_polygon(data = SUB_cristrans_KE_plot, aes(fill = factor(SUB_crisis_trans_count), 
                                             x = long, 
                                             y = lat, 
                                             group = id)) + 
  geom_path(data = SUB_cristrans_KE_plot, aes(x = long, 
                                          y = lat, 
                                          group = group), 
            color = "black", size = 0.1) +
  scale_fill_brewer(name="No of crisis\ntransitions",
                    labels=c("0","1", "2", "3", "4", "5"),
                    palette=("Blues")) +
  coord_equal() +
  #theme_map() + ## switching this one off means you have to delete the longitude & latitude
  theme(legend.position = "bottom") +
  labs(x = "Longitude", 
       y = "Latitude", 
       title = "Crisis transitions per li \n zone of Kenya",
       caption = "Shapefile: SUB_KE, Fill: No Crisis transitions; Data: FEWSNET, 2019")

plot(SUB_Kenya_cristrans_plot)

[See here](https://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example) on making an R question that folks can help with. That includes a sample of data, all necessary code, and a clear explanation of what you're trying to do and what hasn't worked. Without being able to run your code, I'd guess you have an NA level, meaning the issue isn't with ggplot so much as your factor leveling — camille, Feb 04 '20 at 15:02
Can you try to add `drop = FALSE` in `scale_fill_brewer()` and see what happens? — jazzurro, Feb 04 '20 at 15:21
I tried ```drop = FALSE```, but it didn't solve the problem. And I've added the first ten rows of the dataframe i'm using for the plot. The last column is the one for the fill. And @camille you are correct that the dataframe has NA, which it isn't supposed to. I'm checking the code that generated the dataframe to see if I can spot the problem. — Mathijs-ve, Feb 04 '20 at 15:53
I've added the code, which I used to generate the dataframe. I couldn't spot the problem. — Mathijs-ve, Feb 04 '20 at 15:58

score 0 · Accepted Answer · answered Feb 04 '20 at 18:44

There's two problems, so if the factor does not exist at all, even if you use drop=FALSE inside scale_fill_brewer, it will not show up. You have NAs in your counts, and that will be drawn in filling. With the missing factor, it just shifts every into what you see.

Below is a reproducible example:

library(ggplot2)
library(ggmap)
library(maps)
library(mapdata)
library(dplyr)
set.seed(100)
states <- map_data("state")
states=states %>% group_by(region) %>%
mutate(count=sample(c(0,1,2,3,5),1))

# no 4s
table(states$count)
  0    1    2    3    5 
2013 2545 4428 3357 3194 

# no NAs, everything works
ggplot(data = states) + 
  geom_polygon(aes(x = long, y = lat, fill = factor(count), group = group), color = "white") + 
  scale_fill_brewer(
               labels=c("0","1", "2", "3", "4", "5"),
               palette=("Blues"),drop=FALSE)

You see it is quite messed up. Now we introduce one NA:

states$count[states$region %in% c("california","washington","north carolina","michigan")] = NA
ggplot(data = states) + 
  geom_polygon(aes(x = long, y = lat, fill = factor(count), group = group), color = "white") + 
  scale_fill_brewer(labels=c("0","1", "2", "3", "4", "5"),
                    palette=("Blues"))

You get exactly what we see in your plot. The last label is for NAs.

So if we want to keep the NAs, do the factor by including 4 in the levels, and also the "NA" string:

states$count[is.na(states$count)] = "NA"
fillCols = c(RColorBrewer::brewer.pal(6,"Blues"),"#b0a160")
ggplot(data = states) + 
geom_polygon(aes(x = long, y = lat, fill = factor(count,levels=c(0:5,"NA")), 
group = group), color = "white") + 
 scale_fill_manual(name="counts",
                   values=fillCols,drop=FALSE)

you are welcome :) glad it worked for you :) must be quite frustrating. — StupidWolf, Feb 05 '20 at 09:44

ggplot not plotting zeros and missing values

1 Answers1