Question

根据乔恩·斯普林（Jon Spring）的建议，我尝试重新编写以下大部分代码。

library(tidyverse)
ka2 <- ka %>%
  gather(Year, Export, -c(Economy, Partner)) %>%
  group_by(Year) %>%
  arrange(Year, -Export) %>%
  top_n(10, wt = Export) %>%
  ungroup()

ka2$Year <- gsub("X", "", ka2$Year)
ka2$Economy <- NULL

ka2 <- droplevels(ka2)
sapply(ka2, class)
ka2$Year <- as.integer(ka2$Year)

library(ggplot2)
library(scales)


ggplot(ka2, aes(x=reorder(Partner, -Export), y = Export/1000000, fill = Partner)) +
  geom_bar(stat = "identity") +
  scale_y_continuous(labels = comma) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1), 
        legend.position = "none") +
  labs(title = "Kazhakhstan Exports to Largest Partners, 2000-2015", 
       y = "Bln USD", x = element_blank()) +
  facet_wrap(~ Year, scales = 'free_x')

适当时生成的data.frame如下：

    structure(list(Partner = c("Switzerland", "Italy", "Russia", 
"China", "France", "Iran", "Netherlands", "Israel", "Azerbaijan", 
"Spain", "Switzerland", "Italy", "Russia", "France", "China", 
"Iran", "Netherlands", "USA", "Israel", "Canada", "Italy", "Switzerland", 
"Russia", "China", "France", "Iran", "Netherlands", "UK", "Spain", 
"Romania", "Italy", "Switzerland", "China", "Russia", "France", 
"Netherlands", "Iran", "UK", "Ukraine", "Israel", "Italy", "Switzerland", 
"China", "Russia", "France", "Netherlands", "Israel", "Iran", 
"Ukraine", "Turkey", "Italy", "China", "Russia", "France", "Switzerland", 
"Netherlands", "Canada", "Ukraine", "Iran", "UK"), Year = c(2004L, 
2004L, 2004L, 2004L, 2004L, 2004L, 2004L, 2004L, 2004L, 2004L, 
2005L, 2005L, 2005L, 2005L, 2005L, 2005L, 2005L, 2005L, 2005L, 
2005L, 2006L, 2006L, 2006L, 2006L, 2006L, 2006L, 2006L, 2006L, 
2006L, 2006L, 2007L, 2007L, 2007L, 2007L, 2007L, 2007L, 2007L, 
2007L, 2007L, 2007L, 2008L, 2008L, 2008L, 2008L, 2008L, 2008L, 
2008L, 2008L, 2008L, 2008L, 2009L, 2009L, 2009L, 2009L, 2009L, 
2009L, 2009L, 2009L, 2009L, 2009L), Export = c(3760396L, 3108975L, 
2836286L, 1966911L, 1468224L, 712011L, 464600L, 322423L, 287125L, 
281365L, 5509511L, 4190531L, 2926578L, 2665146L, 2422507L, 886118L, 
877836L, 666028L, 661641L, 528132L, 6891644L, 6721180L, 3730037L, 
3592514L, 3346969L, 2077598L, 1704555L, 1143876L, 968365L, 747116L, 
7774224L, 7475877L, 5635914L, 4658919L, 3982705L, 2464262L, 2451368L, 
1133234L, 1113097L, 1058817L, 11920317L, 11281326L, 7676609L, 
6227049L, 5388682L, 4638669L, 2226504L, 2039530L, 2003343L, 1903764L, 
6686756L, 5888593L, 3546967L, 3381509L, 2668219L, 2222452L, 1385352L, 
1289161L, 1279004L, 1235083L)), .Names = c("Partner", "Year", 
"Export"), row.names = c(NA, -60L), class = c("tbl_df", "tbl", 
"data.frame"))

然后，我得到的图像就是这样。

The result

但是，您可以看到，即使每年显示的是前十大出口目的地，它们也没有按照降序排列。 data.frame是通过按降序排列值来保护的，但显示并非如此。希望这可以解决。

根据Uwe的链接和建议，我进一步完善了代码，如下所示。

ka2$ord <- rep(10:1,len=120)

ggplot(ka2, aes(x = -ord, y = Export/1000000, fill = Partner)) +
  geom_bar(stat = "identity") +
  scale_y_continuous(labels = comma) +
  scale_x_discrete(labels = ka2[, setNames(as.character("Partner"), "ord")]) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1), 
        legend.position = "none") +
  labs(title = "Kazhakhstan Exports to Largest Partners, 2004-2015", 
       y = "Bln USD", x = "Partner") +
  facet_wrap(~ Year, scales = "free_x")

它给我的result2如下。

The result 2

问题出在axis.x标签上。

Answer 1

使用基数R，我们可以选择要查找其前10个条目的列（在本例中为3:5），然后对于每个列，我们将order与{{1 }}以降序获取条目并仅选择前10个条目。然后，我们使用这些条目作为索引来创建一个数据框，并从decreasing = TRUE列中将其作为子集，以获取国家/地区，最后使用Partner来获取所有数据框列表cbind

do.call

Answer 2

这是使用public class Tester { public static void main(String[] args) { LinkedList ll = new LinkedList(); ll.push(35); ll.push(100); ll.push(14); ll.push(44); ll.push(10); ll.push(8); System.out.println("Created Linked list is:"); ll.printList(); System.out.println(ll.getSize()); } }和tidyverse的一种方法。

第一部分很简单，将广泛的数据转换成一个长长的列表，并收集每年的前十名。

purrr

要使输出具有重复标题的宽格式，我首先创建一个函数，将年份作为输入，并输出该年份的国家和值。然后，我每年每年使用library(tidyverse) df2_long <- df %>% select(-Economy) %>% gather(year, value, X2000:X2002) %>% group_by(year) %>% arrange(year, -value) %>% top_n(10, wt = value) %>% ungroup()运行该功能，然后将它们绑定在一起。

purrr::map

输出

library(purrr)
year_cols <- unique(df2_long$year)
grab_chunk <- function(year_col) {
  df2_long %>%
  filter(year == year_col) %>%
  mutate(Partner = fct_reorder(Partner, -value)) %>%  # Added to keep order
  spread(year, value)
}
df2_wide <- map(year_cols, grab_chunk) %>%
  bind_cols()

Answer 3

注意：以下答案是指OP完全更改问题之前的原始问题：

原始问题要求将Partner的排名按每年的降序排列，并以广泛的格式显示排名前10位的合作伙伴以及每年的相应值。

为了完整起见，这也是使用melt()和dcast()的答案。该解决方案可以灵活地与任意数量的年份列X2000，X2001等一起使用。

library(data.table)
top <- 10L
val_cols <- c("Partner", "value")
# reshape from wide to long format
long <- melt(setDT(ka), id.vars = c("Economy", "Partner"), variable.name = "year")
# order by decreasing value and append rank for each year
long[order(-value), rank := rowid(year)]
# reshape from long to wide format and keep only the top rows, remove rank
wide <- dcast(long, Economy + rank ~ year, value.var = val_cols)[, rank := NULL][1:top]
# reorder columns for convenience
setcolorder(wide, c("Economy",
            as.vector(outer(val_cols, unique(long$year), paste, sep = "_"))))
wide

       Economy Partner_X2000 value_X2000 Partner_X2001 value_X2001 Partner_X2002 value_X2002
 1: Kazakhstan        Russia     1710262        Russia     1733412        Russia     1497738
 2: Kazakhstan         Italy      917604         Italy      956196         China     1018680
 3: Kazakhstan         China      672549         China      646651         Italy      904222
 4: Kazakhstan       Germany      546887       Germany      495533   Switzerland      773596
 5: Kazakhstan   Switzerland      463171       Ukraine      490215           UAE      472277
 6: Kazakhstan       Ukraine      254241   Switzerland      378243        Poland      320482
 7: Kazakhstan   Netherlands      219459           UAE      330340          Iran      309882
 8: Kazakhstan            UK      219420            UK      271586       Ukraine      291469
 9: Kazakhstan          Iran      203270          Iran      208925       Germany      219732
10: Kazakhstan           USA      176198        Poland      164157            UK      131824

编辑： rank列已根据OP的请求从结果中删除。

附录

OP发布了一个相关问题reshaping data frame with tidyverse，因为ggplot2需要长格式的绘图数据。

此答案中的中间结果long可用于绘制：

long[rank <= 10]

完整的答案：

library(ggplot2)
ggplot(
  long[, ord := sprintf("%03i", frank(long, year, -value, ties.method = "first"))][rank <= 10], 
  aes(x = ord, y = value/1000000, fill = Partner)) +
  geom_col() +
  scale_y_continuous(labels = scales::comma) +
  theme(axis.text.x = element_text(angle = 90, hjust = 1), 
        legend.position = "none") +
  labs(title = "Kazhakhstan Exports to Largest Partners, 2000-2015", 
       y = "Bln USD", x = element_blank()) +
  facet_wrap(~ year, scales = 'free_x', drop = TRUE) + 
  scale_x_discrete(labels = long[, setNames(as.character(Partner), ord)])

数据

以下是OP在原始问题中提供的数据：

ka <- structure(list(Economy = structure(c(1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 1L, 
1L, 1L, 1L, 1L), .Label = c("Kazakhstan", "Kyrgyzstan", "Tajikistan", 
"Turkmenistan", "Uzbekistan"), class = "factor"), Partner = structure(c(1L, 
2L, 3L, 4L, 5L, 6L, 7L, 8L, 9L, 10L, 11L, 12L, 13L, 14L, 16L, 
17L, 18L, 19L, 20L, 31L, 21L, 22L, 23L, 25L, 26L, 27L, 28L, 24L, 
29L, 30L, 32L, 33L, 35L, 36L, 37L, 34L, 40L, 38L, 39L, 41L, 42L, 
15L), .Label = c("Austria", "Azerbaijan", "Bangladesh", "Belarus", 
"Belgium", "Canada", "China", "Czechia", "France", "Georgia", 
"Germany", "Greece", "Hungary", "India", "Indonesia", "Iran", 
"Israel", "Italy", "Japan", "Kazakhstan", "Kyrgyzstan", "Malaysia", 
"Mexico", "Moldova", "Mongolia", "Netherlands", "Pakistan", "Poland", 
"Romania", "Russia", "SouthKorea", "Spain", "Switzerland", "Tajikistan", 
"Thailand", "Turkey", "Turkmenistan", "UAE", "UK", "Ukraine", 
"USA", "Uzbekistan"), class = "factor"), X2000 = c(556L, 46816L, 
2839L, 16155L, 7584L, 4393L, 672549L, 7740L, 4319L, 7646L, 546887L, 
473L, 2717L, 29497L, 203270L, 7033L, 917604L, 9733L, NA, 36430L, 
57247L, 10005L, 6342L, 12240L, 219459L, 1455L, 56626L, 1076L, 
523L, 1710262L, 6514L, 463171L, 15174L, 62280L, 7100L, 52602L, 
254241L, 11008L, 219420L, 176198L, 133526L, NA), X2001 = c(3170L, 
69309L, 3111L, 5120L, 3627L, 333L, 646651L, 4167L, 7217L, 2954L, 
495533L, 4381L, 7904L, 9279L, 208925L, 3599L, 956196L, 15474L, 
NA, 43400L, 83778L, 1374L, NA, 9768L, 141327L, 489L, 164157L, 
2709L, 7837L, 1733412L, 6782L, 378243L, 3153L, 74231L, 14178L, 
61198L, 490215L, 330340L, 271586L, 142412L, 150234L, NA), X2002 = c(3307L, 
112657L, 1170L, 11815L, 7384L, 9492L, 1018680L, 15064L, 27434L, 
5238L, 219732L, 30678L, 12626L, 4213L, 309882L, 5839L, 904222L, 
22998L, NA, 48866L, 108462L, 2624L, NA, 8329L, 123553L, 438L, 
320482L, 2734L, 121044L, 1497738L, 11549L, 773596L, 17849L, 97449L, 
15254L, 45740L, 291469L, 472277L, 131824L, 116942L, 101022L, 
NA)), row.names = c(NA, -42L), class = "data.frame")

ka

      Economy      Partner   X2000   X2001   X2002
1  Kazakhstan      Austria     556    3170    3307
2  Kazakhstan   Azerbaijan   46816   69309  112657
3  Kazakhstan   Bangladesh    2839    3111    1170
4  Kazakhstan      Belarus   16155    5120   11815
5  Kazakhstan      Belgium    7584    3627    7384
6  Kazakhstan       Canada    4393     333    9492
7  Kazakhstan        China  672549  646651 1018680
8  Kazakhstan      Czechia    7740    4167   15064
9  Kazakhstan       France    4319    7217   27434
10 Kazakhstan      Georgia    7646    2954    5238
11 Kazakhstan      Germany  546887  495533  219732
12 Kazakhstan       Greece     473    4381   30678
13 Kazakhstan      Hungary    2717    7904   12626
14 Kazakhstan        India   29497    9279    4213
15 Kazakhstan         Iran  203270  208925  309882
16 Kazakhstan       Israel    7033    3599    5839
17 Kazakhstan        Italy  917604  956196  904222
18 Kazakhstan        Japan    9733   15474   22998
19 Kazakhstan   Kazakhstan      NA      NA      NA
20 Kazakhstan   SouthKorea   36430   43400   48866
21 Kazakhstan   Kyrgyzstan   57247   83778  108462
22 Kazakhstan     Malaysia   10005    1374    2624
23 Kazakhstan       Mexico    6342      NA      NA
24 Kazakhstan     Mongolia   12240    9768    8329
25 Kazakhstan  Netherlands  219459  141327  123553
26 Kazakhstan     Pakistan    1455     489     438
27 Kazakhstan       Poland   56626  164157  320482
28 Kazakhstan      Moldova    1076    2709    2734
29 Kazakhstan      Romania     523    7837  121044
30 Kazakhstan       Russia 1710262 1733412 1497738
31 Kazakhstan        Spain    6514    6782   11549
32 Kazakhstan  Switzerland  463171  378243  773596
33 Kazakhstan     Thailand   15174    3153   17849
34 Kazakhstan       Turkey   62280   74231   97449
35 Kazakhstan Turkmenistan    7100   14178   15254
36 Kazakhstan   Tajikistan   52602   61198   45740
37 Kazakhstan      Ukraine  254241  490215  291469
38 Kazakhstan          UAE   11008  330340  472277
39 Kazakhstan           UK  219420  271586  131824
40 Kazakhstan          USA  176198  142412  116942
41 Kazakhstan   Uzbekistan  133526  150234  101022
42 Kazakhstan    Indonesia      NA      NA      NA

从data.frame的每一列中选取10个最大值，然后以降序排列

3 个答案:

附录

数据