ggplot不绘制零和缺失值答案

【问题标题】：ggplot not plotting zeros and missing valuesggplot不绘制零和缺失值
【发布时间】：2020-02-05 02:44:31
【问题描述】：

我正在尝试制作一张肯尼亚地图，其填充包含值“0-3”和“5”。但是，ggplot 将填充为“0”的县绘制为空白，并在比例尺上移动值，因为没有值“4”。我怎样才能解决这个问题？以绘制零点和比例保持正确的方式。下面是我用来制作情节的代码。还有我用来为绘图制作数据框的代码。

如果您希望我添加一些内容，请告诉我，并提前致谢！

这些是我的数据框的前 10 行：

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)

【问题讨论】：

See here 提出一个人们可以帮助解决的 R 问题。这包括一个数据样本、所有必要的代码，以及对你正在尝试做什么和什么没有奏效的清晰解释。如果无法运行您的代码，我猜您有一个 NA 级别，这意味着问题不在于 ggplot，而在于您的因子水平
您可以尝试在scale_fill_brewer() 中添加drop = FALSE 看看会发生什么？
我试过drop = FALSE，但没有解决问题。我已经添加了用于绘图的数据框的前十行。最后一列是用于填充的列。 @camille 你是正确的，数据框有 NA，这是不应该的。我正在检查生成数据帧的代码，看看是否能发现问题。
我已经添加了用于生成数据帧的代码。我没发现问题。

标签： r ggplot2 spatial

【解决方案1】：

有两个问题，所以如果这个因子根本不存在，即使你在 scale_fill_brewer 中使用drop=FALSE，它也不会出现。您的计数中有 NA，这将在填充时绘制。由于缺少因素，它只是将每一个转换为您所看到的。

下面是一个可重现的例子：

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)

你看它很混乱。现在我们介绍一个 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"))

你得到的正是我们在你的情节中看到的。最后一个标签用于 NA。

因此，如果我们想保留 NA，请通过在级别中包含 4 以及“NA”字符串来执行该因子：

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)

【讨论】：

不客气 :) 很高兴它对你有用 :) 一定很令人沮丧。