问题

我有一个包含德国5位邮政编码的形状文件（shapefile）。德国的大部分1位数字邮政编码与德国各州相似。我使用rgdal读取了这个形状文件数据，因此有了一个SpatialPolygonsDataFrame。我只有德国的某些部分的数据，也就是一些邮政编码。我想在5位数字级别上展示我有的数据。使用leaflet创建地图时，对我来说绘制近10,000个邮政编码需要很长时间。因此，我想要在那些没有数据的邮政编码外边界进行“汇总”/“合并”，其中数值为NA。

# 德国邮政编码形状

# 创建临时文件
temp <- tempfile()
temp2 <- tempfile()

# 下载zip文件并保存到'temp'中
URL <- "https://downloads.suche-postleitzahl.org/v2/public/plz-5stellig.shp.zip"
download.file(URL, temp)

# 解压缩临时文件的内容并保存解压后的内容到'temp2'中
unzip(zipfile = temp, exdir = temp2)

# 读取形状文件
library(rgdal)
GER_postcode <- readOGR(temp2)

head(GER_postcode@data$note)

# 创建子样本
library(tidyverse)

GER_postcode@data$einwohner2 <- ifelse(substr(GER_postcode@data$plz, 1, 1) %in% c("0", "1", "7"), GER_postcode@data$einwohner, NA)

# 绘制子样本
library(leaflet)

qpal <- colorBin("Reds", GER_postcode@data$einwohner2, bins = 10)

leaflet(GER_postcode) %>%
  addPolygons(stroke = TRUE, opacity = 1, fillOpacity = 0.5, smoothFactor = 0.5,
              color = "black", fillColor = ~qpal(einwohner2), weight = 1) %>%
  addLegend(values = ~einwohner2, pal = qpal, title = "Population")

如何使地图显示那些具有值的邮政编码形状并将所有其他值为NA的邮政编码合并起来？

我看了一下library(rgeos)和gUnaryUnion()，它们可以将形状文件中的所有单位合并到外边界。尽管我只需要在子集上执行此操作。

<details>
<summary>英文:</summary>

 I have a 5-digit postcode shapefile for Germany. Big number 1-digits postcodes are similar to German states. I read shapefile data with `rgdal` thus having a `SpatialPolygonsDataFrame`. I have only data for some part of Germany, i.e. some postcodes. The data I have I like to show on a granular 5-digit level. Using `leaflet` to create a map, it makes very long time for me, to plot all of the almost 10.000 postcodes. Thus I like to &quot;summarize&quot;/&quot;combine&quot;/&quot;merge&quot; the outer border of those postcodes where I don&#39;t have data (where the value is `NA`). 

    # German postcode shapes 
    
    # Create temp files
    temp &lt;- tempfile()
    temp2 &lt;- tempfile()
    
    # Download the zip file and save to &#39;temp&#39; 
    URL &lt;- &quot;https://downloads.suche-postleitzahl.org/v2/public/plz-5stellig.shp.zip&quot;
    download.file(URL, temp)
    
    # Unzip the contents of the temp and save unzipped content in &#39;temp2&#39;
    unzip(zipfile = temp, exdir = temp2)
    
    # Read shape file 
    library(rgdal)
    GER_postcode &lt;- readOGR(temp2)
    
    head(GER_postcode@data$note)
    
    # Create subsample 
    library(tidyverse)
    
    GER_postcode@data$einwohner2 &lt;- ifelse(substr(GER_postcode@data$plz, 1, 1) %in% c(&quot;0&quot;, &quot;1&quot;, &quot;7&quot;), GER_postcode@data$einwohner, NA)
    
    # Plot Subsample 
    library(leaflet)
    
    qpal &lt;- colorBin(&quot;Reds&quot;, GER_postcode@data$einwohner2, bins=10)
    
    leaflet(GER_postcode) %&gt;%
      addPolygons(stroke = TRUE,opacity = 1,fillOpacity = 0.5, smoothFactor = 0.5,
                  color=&quot;black&quot;,fillColor = ~qpal(einwohner2),weight = 1) %&gt;%
      addLegend(values=~einwohner2,pal=qpal,title=&quot;Population&quot;)

How can I make the map show those postcode shapes with values and combine all other where the value is `NA`? 

[![enter image description here][1]][1]

I was looking at `library(rgeos)` and `gUnaryUnion()` that units all units in a shapefile to outer borders. Although I only need to do this on a subset.

  [1]: https://i.stack.imgur.com/Tkcx7.png

</details>


# 答案1
**得分**: 2

我个人会使用过滤器/子集来删除不相关的多边形，并将国家轮廓用作单独的背景层。轮廓可以通过将邮政编码形状联合起来来构建，虽然下载一个（或安装一些捆绑国家形状的软件包）可能更快。

由于“rgdal”和“regos”将在几个月内停用，下面的示例使用“sf”（以及可选的“geos”软件包）。

```r
library(sf)
#&gt; Linking to GEOS 3.9.3, GDAL 3.5.2, PROJ 8.2.1; sf_use_s2() is TRUE
library(dplyr)
library(leaflet)

URL &lt;- "https://downloads.suche-postleitzahl.org/v2/public/plz-5stellig.shp.zip"

# 使用GDAL虚拟文件系统从远程URL加载压缩的shapefile
GER_postcode &lt;- paste0("/vsizip//vsicurl/", URL) %&gt;%  read_sf()

# 从giscoR获取国家轮廓
GER_outline &lt;- giscoR::gisco_get_countries(country = "DE")

# 如果必须从输入形状构建它，我们可以使用sf::st_union()或切换到geos（R库），它在这种情况下表现得更好：

# GER_outline &lt;- as_geos_geometry(GER_postcode) %&gt;% 
#   geos_make_collection() %&gt;% 
#   geos_unary_union() %&gt;% 
#   st_as_sfc()

GER_postcode_subsample &lt;- GER_postcode %&gt;% filter(substr(plz, 1, 1) %in% c(0, 1, 7))

# 绘制子样本
qpal &lt;- colorBin("Reds", GER_postcode_subsample$einwohner, bins=10)

leaflet(GER_postcode_subsample) %&gt;%
  addPolygons(data = GER_outline, smoothFactor = 0.5,
              color="black", fillColor = "#808080", weight = 1) %&gt;% 
  addPolygons(stroke = TRUE,opacity = 1,fillOpacity = 0.5, smoothFactor = 0.5,
              color="black",fillColor = ~qpal(einwohner),weight = .2) %&gt;%
  # 将NA附加到值以包含在图例中
  addLegend(values=c(~einwohner, NA),pal=qpal,title="Population")

<!-- -->

^{创建于2023年7月17日，使用reprex v2.0.2}

library(sf)
#> Linking to GEOS 3.9.3, GDAL 3.5.2, PROJ 8.2.1; sf_use_s2() is TRUE
library(dplyr)
library(leaflet)

URL <- "https://downloads.suche-postleitzahl.org/v2/public/plz-5stellig.shp.zip"

# use GDAL virtual file systems to load zipped shapefile from remote url
GER_postcode <- paste0("/vsizip//vsicurl/", URL) %>%  read_sf()

# country outline from giscoR
GER_outline <- giscoR::gisco_get_countries(country = "DE")

# if it must be built from input shapes, we can use sf::st_union() or switch  
# to geos (the R library), it performs bit better in this case:

# GER_outline <- as_geos_geometry(GER_postcode) %>% 
#   geos_make_collection() %>% 
#   geos_unary_union() %>% 
#   st_as_sfc()

GER_postcode_subsample <- GER_postcode %>% filter(substr(plz, 1, 1) %in% c(0, 1, 7))

# Plot Subsample 
qpal <- colorBin("Reds", GER_postcode_subsample$einwohner, bins=10)

leaflet(GER_postcode_subsample) %>%
  addPolygons(data = GER_outline, smoothFactor = 0.5,
              color="black", fillColor = "#808080", weight = 1) %>% 
  addPolygons(stroke = TRUE,opacity = 1,fillOpacity = 0.5, smoothFactor = 0.5,
              color="black",fillColor = ~qpal(einwohner),weight = .2) %>%
  # append NA to values to include it in legend
  addLegend(values=c(~einwohner, NA),pal=qpal,title="Population")