數(shù)據(jù)可視化總是看著簡單，但實(shí)操起來bug頻繁，究其原因，還是數(shù)據(jù)處理和ggplot2的相關(guān)參數(shù)等掌握不精。

1. 數(shù)據(jù)準(zhǔn)備

#加載包
library(dplyr)
library(readr)
library(ggplot2)

#讀取數(shù)據(jù)
athlete_events <- read_csv("athlete_events.csv")
noc_regions <- read_csv("noc_regions.csv")
head(athlete_events)

athletedata <- inner_join(athlete_events,noc_regions,c("NOC"))
head(athletedata)

2. 了解每個地區(qū)參數(shù)人數(shù)的分布情況(條形圖)

region <- athletedata %>% 
    group_by(region) %>% 
    summarise(value=n()) %>% 
    arrange(desc(value)) #按照地區(qū)分組，計算每個地區(qū)的參賽人數(shù)，并按照從大到小的順序輸出

region30 <- region[1:30,]
region_plot <- ggplot(region30,aes(x=reorder(region,value),y=value)) +
    theme_bw(base_family = "STKaiti") +
    geom_bar(aes(fill=value),stat = "identity",show.legend = F) +
    coord_flip() +
    scale_fill_gradient(low = "blue",high = "red") +
    labs(x="地區(qū)",y="參賽人數(shù)",title = "每個地區(qū)參賽人數(shù)的分布情況") +
    theme(axis.text.x = element_text(vjust = 0.5),plot.title = element_text(hjust = 0.5))
region_plot

條形圖

3. 了解不同時間、不同地區(qū)、不同性別的參賽人數(shù)的分布情況(熱力圖)

index <- region30$region[1:30] #取前30個地區(qū)為一組向量，而region30[1:30,1]仍為列表形式

region30_merge![](http://upload-images.jianshu.io/upload_images/25726527-67da2172b54802b3.jpg?imageMogr2/auto-orient/strip%7CimageView2/2/w/1080/q/50)

 <- athletedata %>% 
    filter(region %in% index) %>%
    group_by(Year,region,Sex) %>%
    summarise(value=n())

merge_plot <- ggplot(region30_merge,aes(x=Year,y=region)) +
    theme_bw(base_family = "STKaiti") +
    geom_tile(aes(fill=value),color="white") +
    scale_fill_gradientn(colors = c("blue","red")) +
    scale_x_continuous(breaks = unique(region30_merge$Year)) +
    theme(axis.text.x = element_text(hjust = 0.5,angle = 90)) +
    facet_wrap(~Sex,nrow = 2)
options(repr.plot.width=10, repr.plot.height=8)
merge_plot

熱力圖

4. 將USA、Germany、France、UK、Russia、China這6個地區(qū)每年奧運(yùn)會獎牌的數(shù)量可視化(折線圖)

index <- c("USA","Germany","France","UK","Russia","China")
region6 <- athletedata %>% 
    filter(region %in% index) %>% #只選擇6個地區(qū)
    filter(Medal!="NA") %>% #只選擇有獎牌獲得
    group_by(region,Year) %>% 
    summarise(value=n())

region6_plot <- ggplot(region6,aes(x=Year,y=value)) +
    theme_bw(base_family = "STKaiti") +
    geom_line() +
    facet_wrap(~region,nrow = 3)
region6_plot

折線圖

5. 動態(tài)展示不同地區(qū)每年的獎牌獲得情況

library(gganimate)
index <- region30$region[1:30]
region30_medal <- athletedata %>% 
    filter(region %in% index) %>% 
    filter(Medal!="NA") %>% 
    group_by(region,Year) %>% 
    summarise(value=n())
region30_medal$Year <- as.integer(region30_medal$Year) #將year這一變量變?yōu)檎妥兞?
region30_plot <- ggplot(region30_medal,aes(x=region,y = value)) +
    theme_bw() +
    geom_bar(stat = "identity",show.legend = F) +
    theme(axis.text.x = element_text(hjust = 0.5,angle = 90)) +
    transition_time(Year) +
    labs(title="Year: {frame_time}")
region30_plot

6. 對地區(qū)、運(yùn)動員數(shù)量、性別、獎牌數(shù)量這些變量進(jìn)行可視化分析(樹圖)

library(treemap)
#計算獎牌數(shù)量
medal <- athletedata %>% 
    filter(Medal!="NA") %>% 
    group_by(region,Sex) %>% 
    summarise(medalnum=n())

#計算運(yùn)動員數(shù)量
athelete <- athletedata %>% 
    group_by(region,Sex) %>% 
    summarise(atheletenum=n())
data <- inner_join(medal,athelete,c("region","Sex"))

data_plot <- treemap(data,index = c("Sex","region"),vSize = "atheletenum",
                    vColor="medalnum",type="value",palette=c("blue","red"),
                    title="不同性別下不同地區(qū)運(yùn)動員的數(shù)量",title.legend="獎牌數(shù)量",
                    fontfamily.title="STKaiti",fontfamily.legend="STKaiti")
data_plot

樹圖

日常廢話：很多參數(shù)的設(shè)置都可以理解為圖層的疊加，因此只要心中有大致的圖形輪廓，圍繞這個輪廓進(jìn)行添枝加葉，就能更加靈活應(yīng)用各種參數(shù)?？赡苊總€新手都會同我一樣，面對這些措手不及，甚至產(chǎn)生畏懼心理，覺得自己可能就不是學(xué)代碼的這塊料，自己就是不行，學(xué)啥都慢，等等。有這些想法都是正常的，因為很多知識都是我們需要用了才會去了解去掌握，為了進(jìn)步為了更加優(yōu)秀才會深入拓展，所以一定要給自己運(yùn)用這些知識的機(jī)會，在實(shí)踐中你才能享受到解決問題的快樂?。?！