Rで折れ線グラフ、棒グラフ (新型コロナウイルス：Coronavirus)

ちょっと工夫した折れ線グラフ、棒グラフを作りました。

新型コロナウイルスの感染状況

米ジョンズ・ホプキンス大学の新型コロナウイルスの感染状況をまとめたWebサイト
Coronavirus 2019-nCoV Global Cases by Johns Hopkins CSSE

データはGitHubから入手できます。
2019 Novel Coronavirus COVID-19 (2019-nCoV) Data Repository by Johns Hopkins CSSE

使用するデータ（3/22からファイル場所、名前変更）
CSSE COVID-19 Dataset

新型コロナウイルスに感染された方、回復された方、亡くなった方の数の推移（日別）

y軸のラベルの数値にコンマをつけ、x軸のラベルを選択して表示

致死率(%):Deaths/Confirmed の推移

(%)の表示をy軸の上部に配置、x軸のラベルを選択して22 Jan.のように表示

死亡者数900人以上の国の死亡者数(表)

	7/2/20
US	128740
Brazil	61884
United Kingdom	44080
Italy	34818
France	29878
Mexico	29189
Spain	28368
India	18213
Iran	11106
Peru	10045
Belgium	9761
Russia	9668
Germany	9006
Canada	8700
Netherlands	6132
Chile	5920
Sweden	5411
Turkey	5167
China	4641
Ecuador	4639
Pakistan	4551
Colombia	3650
Egypt	3120
Indonesia	2987
South Africa	2844
Iraq	2160
Switzerland	1965
Bangladesh	1926
Saudi Arabia	1752
Ireland	1738
Romania	1687
Portugal	1587
Poland	1492
Argentina	1385
Philippines	1274
Bolivia	1271
Ukraine	1200
Japan	977
Algeria	928

死亡者数900人以上の国の致死率(%):Deaths/Confirmed の推移

報告された感染者数が一定数超えた時点を起点とした折れ線グラフ

片対数グラフ（Semi-log plot）

(参考)Alessandro Strumia(physicist) Twitter:Days since reported cases reach 200

Days since reported cases reach 600

報告された感染者数が10000人以上の国ごとの感染者数の棒グラフ

x軸のラベルの数値にコンマ、国ごとの値をグラフ右に表記

報告された感染者数が10000人以上の国ごとの致死率の棒グラフ

日本と韓国のグラフの色、ラベルの色を変えた。

報告された感染者数 / 人口 * 100 (タイトルにbquote関数を使う)

country	pop	population_density	Confirmed	Confirmed_per_Pop
China	1355692576	141.26	84830	0.006257
Japan	127103388	336.33	19055	0.014992
Nigeria	177155754	191.78	27110	0.015303
Indonesia	253609643	133.16	59394	0.023419
Korea, South	49039986	491.78	12967	0.026442
Philippines	107668231	358.89	38805	0.036041
Algeria	38813722	16.30	14657	0.037762
Morocco	32987206	73.87	12969	0.039315
Nepal	30986975	210.54	14519	0.046855
India	1236344631	376.10	625544	0.050596
Cameroon	23130708	48.65	12592	0.054438
Ghana	25758108	107.99	18134	0.070401
Egypt	86895099	86.77	71299	0.082052
Poland	38346279	122.64	35146	0.091654
Bangladesh	166280712	1154.74	153277	0.092180
Afghanistan	31822848	48.79	32022	0.100626
Ukraine	44291413	73.38	46821	0.105711
Pakistan	196174380	246.42	221896	0.113112
Czechia	10627448	134.75	12178	0.114590
Romania	21729871	91.15	27746	0.127686
Guatemala	14647083	134.51	20072	0.137038
Argentina	43024374	15.47	69941	0.162561
Iraq	32585692	74.34	53708	0.164821
Azerbaijan	9686210	111.85	18684	0.192893
Mexico	120286655	61.23	238511	0.198286
Serbia	7209764	93.06	15195	0.210756
Austria	8223062	98.04	17941	0.218179
Colombia	46245297	40.60	102261	0.221127
Denmark	5569077	129.23	13015	0.233701
Kazakhstan	17948816	6.59	42574	0.237197
Germany	80996685	226.87	196370	0.242442
Honduras	8598561	76.71	21120	0.245622
Turkey	81619392	104.16	202284	0.247838
Iran	80840713	49.05	232863	0.288052
Netherlands	16877351	406.26	50546	0.299490
Canada	34834841	3.49	106643	0.306139
France	66259012	102.92	203640	0.307339
Dominican Republic	10349741	212.65	34197	0.330414
Bolivia	10631486	9.68	35528	0.334177
Israel	7821850	376.59	27047	0.345788
South Africa	48375645	39.68	168061	0.347408
Ecuador	15654411	55.21	59468	0.379880
Italy	61680122	204.69	240961	0.390662
Portugal	10813834	117.43	42782	0.395623
Switzerland	8061516	195.30	31967	0.396538
United Kingdom	63742977	261.66	285285	0.447555
Russia	142470272	8.33	660231	0.463417
Moldova	3583288	105.85	17150	0.478611
Spain	47737941	94.46	250103	0.523908
Ireland	4832765	68.77	25489	0.527421
Belgium	10449361	342.29	61598	0.589491
Belarus	9608058	46.28	62698	0.652556
Saudi Arabia	27345986	12.72	197608	0.722622
Sweden	9723809	21.59	70639	0.726454
Brazil	202656788	23.80	1496858	0.738617
Singapore	5567301	7987.52	44310	0.795897
US	318892103	32.45	2741180	0.859595
Armenia	3060631	102.90	26658	0.870997
United Arab Emirates	5628805	67.33	49469	0.878854
Peru	30147935	23.46	292004	0.968570
Panama	3608431	47.84	35237	0.976519
Oman	3219775	10.40	42555	1.321676
Chile	17363894	22.97	284541	1.638693
Kuwait	2742711	153.93	47859	1.744952
Bahrain	1314089	1729.06	27837	2.118350
Qatar	2123160	183.25	97897	4.610910

日本の人口あたり（報告された！！）感染者が少ないのは検査しないから。

R code : 韓国と日本のPCR検査実施人数比較 (新型コロナウイルス：Coronavirus)

Plot in Plot

Rコード

データをGitHubから入手。(read.csvの際には、check.names=Fをつける)

# read.csvの際には、check.names=Fをつける
url<- "https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_confirmed_global.csv"
Confirmed<- read.csv(url,check.names=F)
url<- "https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_deaths_global.csv"
Deaths<- read.csv(url,check.names=F)
url<- "https://raw.githubusercontent.com/CSSEGISandData/COVID-19/master/csse_covid_19_data/csse_covid_19_time_series/time_series_covid19_recovered_global.csv"
Recovered<- read.csv(url,check.names=F)
#save(Confirmed,file="Confirmed.Rdata") ; save(Recovered,file="Recovered.Rdata") ; save(Deaths,file="Deaths.Rdata") 

全世界の日毎の合計：Confirmed , Deaths, Recovered 別

# Confirmed , Deaths, Recoveredのデータ数が違うことがある。
# rbindを使うと不具合あり。もしくはエラーとなる。
nCoV1<-colSums(Confirmed[,5:ncol(Confirmed)])
nCoV2<-colSums(Deaths[,5:ncol(Deaths)])
nCoV3<-colSums(Recovered[,5:ncol(Recovered)])
lenmax<- max(c(length(nCoV1),length(nCoV2),length(nCoV3)))
nCoV<-matrix(NA,nrow=3,ncol=lenmax)
nCoV[1,1:length(nCoV1)]<- nCoV1
nCoV[2,1:length(nCoV2)]<- nCoV2
nCoV[3,1:length(nCoV3)]<- nCoV3
rownames(nCoV)<- c("Confirmed","Deaths","Recovered")
colnames(nCoV)<- names(nCoV1)

感染者、回復された方、亡くなった方の数の推移（日別）

旧

# 指数表示を抑制
options(scipen=2) 
#png("Coronavirus001.png",width=800,height=600)
par(mar=c(5,6,3,6),family="serif")
matplot(t(nCoV),type="o",col=1:3,lwd=1.5,lty=1:3,pch=16:18,las=1,xaxt="n",yaxt="n",xlab="",ylab="",bty="l")
box(bty="l",lwd=2)
#axis(1,at=1:ncol(nCoV), labels =sub("/20","",colnames(nCoV)))
#表示するx軸ラベルを指定
axis(1,at=1:ncol(nCoV), labels =NA,tck= -0.01)
labels<-sub("/20","",colnames(nCoV))
#labelpos<- c("2/1","2/10","2/20","3/1","3/10","3/20")
labelpos<- paste0(rep(1:12,each=3),"/",c(1,10,20))
axis(1,at=1,labels =labels[1],tick=F)
for (i in labelpos){
    at<- match(i,labels)
    if (!is.na(at)){ axis(1,at=at,labels = i,tck= -0.02)}
    }
# Add comma separator to axis labels
axis(side=2, at=axTicks(2), labels=formatC(axTicks(2), format="d", big.mark=','),las=1) 
legend("topleft", legend = rownames(nCoV),col=1:3,lwd=1.5,lty=1:3,pch=16:18,inset =c(0.02,0.03))
text(x=par("usr")[2],y=apply(nCoV,1,max,na.rm=T),xpd=T,pos=3,
    labels=formatC(apply(nCoV,1,max,na.rm=T), format="d", big.mark=','),cex=1.5)
title("Coronavirus [ Total Confirmed,Total Recovered,Total Deaths ]")
#dev.off()

新

# 指数表示を抑制
options(scipen=2) 
#
XY <- data.frame(Under_Isolation=nCoV[1,]-(nCoV[2,] + nCoV[3,]), Deaths=nCoV[2,], Recovered=nCoV[3,])
rownames(XY)<- sub("/20","",colnames(nCoV))
#png("Coronavirus001.png",width=800,height=600)
par(mar=c(5,6,4,7),family="serif")
b<- barplot(t(XY),col=c(rgb(1,1,0,0.8),rgb(1,0,0,0.8),rgb(0,1,0,0.8)),yaxt="n",ylim=c(0,max(nCoV[1,],na.rm=T)*1.1),xaxt="n",
    legend=T,args.legend=list(x="topleft",inset=0.03))
lines(x=b,y=nCoV[1,],lwd=3)
#表示するx軸ラベルを指定
#axis(1,at=b[1]:b[nrow(XY)], labels =NA,tck= -0.01)
labels<- rownames(XY)
#axis(1,at=b[1],labels =labels[1],tick=F)
labelpos<- paste0(rep(1:12,each=3),"/",1)
for (i in labelpos){
    at<- b[match(i,labels)]
    if (!is.na(at)){ axis(1,at=at,labels = i,tck= -0.02)}
    }
labelpos<- paste0(rep(1:12,each=3),"/",c(10,20))
for (i in labelpos){
    at<- b[match(i,labels)]
    if (!is.na(at)){ axis(1,at=at,labels = i,tck= -0.01)}
    }
# Add comma separator to axis labels
axis(side=2, at=axTicks(2), labels=formatC(axTicks(2), format="d", big.mark=','),las=1) 
text(x=par("usr")[2],y=c(XY[nrow(XY),1]/2,XY[nrow(XY),1]+max(XY[,2],na.rm=T)/2,XY[nrow(XY),1]+max(XY[,2],na.rm=T)+max(XY[,3],na.rm=T)/2),
    labels=formatC(c(XY[nrow(XY),1],max(XY[,2],na.rm=T),max(XY[,3],na.rm=T)),format="d", big.mark=','),xpd=T,pos=4)
text(x=par("usr")[2],y=max(nCoV[1,],na.rm=T),
    labels= paste0("Confirmed\n",formatC(max(nCoV[1,],na.rm=T), format="d", big.mark=',')),col="darkgreen",xpd=T,pos=4)
title("回復された方、亡くなった方、療養中の方の推移(新型コロナウイルス）",cex.main=1.5)
#dev.off()

致死率(%):Deaths/Confirmed の推移

# nCoVの2行目 / nCoVの1行目
MR<- round(nCoV[2,]/nCoV[1,] *100,3)
#png("Coronavirus01_1.png",width=800,height=600)
par(mar=c(5,6,3,4),family="serif")
plot(MR,type="l",lwd=2.5,las=1,xaxt="n",ylab="Reported deaths / Reported cases",bty="l",xlab="")
box(bty="l",lwd=2)
labels<-sub("/20","",colnames(nCoV))
labels<-gsub("^.*/","",labels)
pos<-gsub("/.*$","",sub("/20","",colnames(nCoV)))
axis(1,at=1:ncol(nCoV),labels =NA)
#月の区切り
#axis(1,at=cumsum(as.vector(table(pos)))+0.5, labels =NA,tck=-0.1,lty=2 ,lwd=1)
axis(1,at=1,labels ="22")
for (i in c("1","10","20")){
    at<- grep("TRUE",is.element(labels,i))
    axis(1,at=at,labels = rep(i,length(at)))
    }
#Month<-c("January","February","March","April","May","June","July","August","September","October","November","December")
Month<-c("Jan.","Feb.","Mar.","Apr.","May","Jun.","Jul.","Aug.","Sep.","Oct.","Nov.","Dec.")
#cut(1:12,breaks = seq(0,12),right=T, labels =Month)
mon<-cut(as.numeric(names(table(pos))),breaks = seq(0,12),right=T, labels =Month)
# 月の中央
#mtext(text=mon,at=cumsum(as.vector(table(pos)))-as.vector(table(pos)/2),side=1,line=2) 
# 月のはじめ
mtext(text=mon,at=1+cumsum(as.vector(table(pos)))-as.vector(table(pos)),side=1,line=2) 
text(x=par("usr")[1],y=par("usr")[4],labels="(%)",pos=2,xpd=T,font=2)
text(x=par("usr")[2],y=MR[length(MR)],labels=paste(MR[length(MR)],"%"),xpd=T)
title("Covid-19 : Death rates(致死率)")
#dev.off()

死亡者数　900人以上の国の致死率(%):Deaths/Confirmed の推移

# Country/Regionごとに集計
#Confirmed
Ctl<- aggregate(Confirmed[,5:ncol(Confirmed)], sum, by=list(Confirmed$"Country/Region"))
rownames(Ctl)<-Ctl[,1]
Ctl<- Ctl[,-1]
#Deaths
Dtl<- aggregate(Deaths[,5:ncol(Deaths)], sum, by=list(Deaths$"Country/Region"))
rownames(Dtl)<-Dtl[,1]
Dtl<- Dtl[,-1]
#
# 死亡者数　900人以上の国 
#
min<- 900
max<- Inf
dat<-Dtl[apply(Dtl,1,max,na.rm=T)>=min & apply(Dtl,1,max,na.rm=T)< max,] 
# "Diamond Princess"を除く
dat<-dat[grep("Diamond Princess",rownames(dat),invert = T),]
#rownames(dat)
dat<- dat[order(dat[,ncol(dat)],decreasing=T),]
knitr::kable(dat[,ncol(dat),drop=F])
#
# 致死率(%)計算
#DpC<- matrix(NA,nrow=nrow(dat),ncol=ncol(dat))
DpC<- NULL
for (i in rownames(dat)){
    temp<- round(dat[rownames(dat)== i,] / Ctl[rownames(Ctl)== i,]*100,2)
    DpC<- rbind(DpC,temp)
}
#
DpC<- DpC[order(DpC[,ncol(DpC)],decreasing=T),]
n<-nrow(DpC)
col<- rainbow(n)
pch<-rep(c(0,1,2,4,5,6,15,16,17,18),3)
#png("Coronavirus01_1_1.png",width=800,height=600)
par(mar=c(3,5,4,10),family="serif")
#40日めから
matplot(t(DpC)[40:ncol(DpC),],type="o",lwd=2,pch=pch,las=1,col=col,ylab="Reported Deaths/Reported Confirmed(%)",xaxt="n",ylim=c(0,20))    
axis(1,at=1:nrow(t(DpC)[40:ncol(DpC),]),labels=sub("/20","",rownames(t(DpC)[40:ncol(DpC),])))
legend(x=par("usr")[2],y=par("usr")[4],legend=rownames(DpC),pch=pch,lwd=2,col=col,bty="n",title="Country/Region",xpd=T)
title(bquote("Reported Deaths / Reported Confirmed (%) ( Reported"~Deaths>=.(min)~")"))
#dev.off()

報告された感染者数が一定数超えた時点を起点とした折れ線グラフ

片対数グラフ（Semi-log plot）ここでは、報告された感染者数 10000人以上

• min,max : 報告された感染者数 min 人以上、max 未満
• Sp : Starting pointとする感染者数
• length : Starting pointとする感染者数に合わせるための調整
• xlim : x軸の範囲

min<- 10000
max<- +Inf
G600<- Ctl[apply(Ctl,1,max,na.rm=T)>= min & apply(Ctl,1,max,na.rm=T)< max,] 
# Diamond Princessを除く
G600<-G600[grep("Diamond Princess",rownames(G600),invert =T),]
G600<-G600[order(apply(G600,1,max,na.rm=T),decreasing=T),]
col<- rainbow(nrow(G600))
#
#Starting point
Sp<- 600 # Starting pointとする感染者数
length<- 10 # Starting pointとする感染者数に合わせるための調整
xlim=c(-10,45) # 範囲
col<- rainbow(nrow(G600))
pch<- rep(c(0,1,2,4,5,6,15,16,17,18),6)
#png("CoronavirusG600.png",width=800,height=600)
par(mar=c(5,4,4,2),family="serif")
# 
plot(1,1,type="n",xlab=paste0("Days since reported cases reach ",Sp),
    ylab="Number of reported cases (log10)",yaxt="n",log="y",
    ylim=c(0.9,2*10^5),xlim=xlim,yaxs="i")
abline(v=seq(-20,40,10),col=c("gray","gray","black","gray","gray","gray","gray"),lty=3)
for(i in 0:6){
  abline(h=seq(1,9)*10^i,col="gray",lty=3)
}
box(lwd=2.5)
#
for (i in 1:length(col)){
p0<- as.numeric(G600[i,])
p<-NULL
for (j in 1:(length(p0)-1)){
    p<- c(p,seq(p0[j],p0[j+1],length=length))
}
p1<- length(p[p<Sp])
p2<- length(p[p>=Sp])
lines(seq(-p1/length,(p2/length - 1/length),1/length),p,lwd=0.8,col=col[i])
points(seq(-p1,p2,length)/length,p0,cex=0.8,col=col[i],pch=pch[i])
if (i== grep("Japan",rownames(G600))){
    text(x=p2[length(p2)]/length,y=p0[length(p0)],labels="Japan",cex=1.2,col="black",pos=4)
    }
if (i== grep("Korea, South",rownames(G600))){
    text(x=p2[length(p2)]/length,y=p0[length(p0)],labels="South Korea",cex=1.2,col="black",pos=4)
    }
#if (i== grep("Singapore",rownames(G600))){
#    text(x=p2[length(p2)]/length,y=p0[length(p0)],labels="Singapore",cex=1.2,col="black",pos=4)
#    }
}
for(i in 0:6){
  axis(side=2, at=10^i, labels=bquote(10^.(i)) ,las=1)
  axis(side=2, at=seq(2,9)*10^i, tck=-0.01,labels=F)
}
legend(x="bottomright",inset=c(0.03,0.01),ncol=4,legend=rownames(G600),pch=pch,lwd=1,col=col,xpd=T,
    bty="n",x.intersp= 1,y.intersp =1,cex=1)
#legend(x=par("usr")[2],y=10^par("usr")[4],legend=rownames(G600),pch=pch,lwd=1,col=col,xpd=T,bty="n",cex=1,ncol=2)
# dev.off()

最後の日付のデータのみ取り出す(drop=F : data.frameで)

Confirmed,Deathsのデータをmergeし、致死率計算

# key : row.namesでmerge
DRbyC<- merge(Ctl[,ncol(Ctl),drop=F],Dtl[,ncol(Dtl),drop=F],by="row.names")
#DRbyC$Row.namesをrownames(DRbyC)に
rownames(DRbyC)<- DRbyC$Row.names
DRbyC<- DRbyC[,-1]
colnames(DRbyC)<- c("Confirmed","Deaths")
# death rate = Deaths/Confirmed *100 
DRbyC$Death.rate<- round(DRbyC$Deaths/DRbyC$Confirmed *100,4)

報告された感染者数 10000人以上の国だけをとりだす。

min<- 10000
DR<- DRbyC[DRbyC$Confirmed >= min ,] 
# Diamond Princessを除く
DR<- DR[grep("Diamond Princess",rownames(DR),invert =T),]

報告された感染者数が10000人以上の国ごとの感染者数の棒グラフ

# Confirmedの昇順に並べ替え
DR<- DR[order(DR$Confirmed,decreasing=F),]
#png("Coronavirus01_2.png",width=800,height=800)
par(mar=c(5,10,3,7),family="serif")
b<- barplot(t(DR)[1,],las=1,horiz=T,xaxt="n",col="pink")
axis(side=1, at=axTicks(1), labels=formatC(axTicks(1), format="d", big.mark=','))
text(x=t(DR)[1,],y=b,labels=formatC(t(DR)[1,], format="d", big.mark=','),pos=4,xpd=T,font=1)
title(bquote("Reported confirmed COVID-19 cases (Country/Region : "~Positive>=.(min)~")"))
#dev.off()

報告された感染者数が10000人以上の国ごとの致死率の棒グラフ

# Death.rateの昇順に並べ替え
DR<- DR[order(DR$Death.rate,decreasing=F),]
#
# Japan の色だけ変える
TF<- is.element(colnames(t(DR)),c("Japan"))
col<- gsub("TRUE","red",gsub("FALSE","lightblue",TF))
col2<- gsub("TRUE","red",gsub("FALSE","black",TF))
# 韓国の色も変えたいなら
col[grep("Korea, South",rownames(DR))]<- "orange"
col2[grep("Korea, South",rownames(DR))]<- "orange"
# png("Coronavirus01_3.png",width=800,height=800)
par(mar=c(6,10,4,3),family="serif")
b<- barplot(t(DR)[3,],las=1,col=col,horiz=T,names=NA)
axis(2, at = b,labels=NA,tck= -0.008)
text(x=par("usr")[1],y=b, labels = colnames(t(DR)), col = col2,pos=2,xpd=T,font=3)
mtext("Reported deaths / Reported cases (%)",side=1,line=3)
title(bquote("Covid-19 : Death rates by country (%) (Country/Region : "~Positive>=.(min)~")"))
#dev.off()

githubで見つけたDataComputingパッケージの CountryDataデータ

CountryData: Many variables on countries from the CIA factbook, 2014.

#install.packages("remotes")
#remotes::install_github("DataComputing/DataComputing")
library(DataComputing)
data("CountryData")
#country: Name of the country.
#area: area (sq km), 2147
#pop: number of people
#
# 報告された感染者数 10000人以上の国
min<- 10000
G10000<- Ctl[apply(Ctl,1,max,na.rm=T)>= min ,] 
# Diamond Princessを除く
G10000<-G10000[grep("Diamond Princess",rownames(G10000),invert =T),]
G10000<-G10000[order(apply(G10000,1,max,na.rm=T),decreasing=T),]
cdata<- CountryData[is.element(CountryData$country,rownames(G10000)),1:3]
#nrow(G10000) ; nrow(cdata) が等しくない
rownames(G10000)[!is.element(rownames(G10000),cdata$country)]
#[1] "US"      "Czechia"
#CountryData$country
# アメリカのnameがCountryDataは"United States" ジョンズ・ホプキンス大学のデータは"US"
# チェコ共和国のnameがCountryDataは"Czech Republic"ジョンズ・ホプキンス大学のデータは"Czechia"
( cdata<- CountryData[is.element(CountryData$country,c("Czech Republic","United States",rownames(G10000))),1:3] )
# 確認(同数か否か)
nrow(G10000) ;nrow(cdata)
#
# cdataのUnited StatesをUSに変更(ジョンズ・ホプキンス大学のデータに合わせる)
cdata$country<- sub("United States","US",cdata$country)
cdata$country<- sub("Czech Republic","Czechia",cdata$country)
# 人口密度(population density)
cdata$population_density<- round(cdata[,3]/cdata[,2],2)
# 最新の報告された感染者のみ取り出し、merge
dat<- G10000[,ncol(G10000),drop=F]
dat$country<- rownames(dat)
cdata<- merge(cdata,dat,by="country")
colnames(cdata)[5]<- "Confirmed"
# 報告された感染者 / 人口 *100
cdata$Confirmed_per_Pop<- round(cdata$Confirmed/cdata$pop*100,6)
# 報告された感染者 / 人口 *100 の昇順に並べ替え
cdata<- cdata[order(cdata$Confirmed_per_Pop,decreasing=F),]
knitr::kable(cdata[,c(1,3:6)],row.names=F)

barplot

#png("Coronavirus01_4.png",width=800,height=800)
par(mar=c(5,8,4,2),family="serif")
barplot(cdata$Confirmed_per_Pop,names=cdata$country,las=1,horiz=T,col="lightblue")
title(bquote("Reported confirmed / number of people *100 (Reported "~confirmed>=.(min)~")"))
#dev.off()

Plot in Plot

load("Wtest.Rdata")
load("Jtest.Rdata")
shp <- system.file("shapes/jpn.shp", package = "NipponMap")[1]
m <- sf::read_sf(shp)
for (i in 1:47){
    Jtest[i,1]<- sub(Jtest[i,1],m$name[i],Jtest[i,1])
}
# 人口100万人あたりの検査数
Jtest$人口100万人あたりの検査数<- round((1000000*Jtest$検査人数)/m$population,0)
# 陽性者数 > 20
df2<- Jtest[Jtest$陽性者数>20,c(1,3,5)]
df3<- Wtest[Wtest[,1]=="Japan",c(1,2,6)]
colnames(df2)<- colnames(df3)
df2<- rbind(df3,df2)
#
lim<-max(Wtest[,"Tests /millionpeople"],na.rm=T)*1.1
min<- 10000
df<- Wtest[!is.na(Wtest[,"Positive"]),]
df<- df[df[,"Positive"]>=min,]
jnum<- df[df[,1]=="Japan","Tests /millionpeople"]
#
# Tests /millionで並べ替え
df<- df[order(df[,"Tests /millionpeople"]),]
color<- is.element(df[,1],"Japan")
col<- gsub("FALSE","lightblue",gsub("TRUE","red",color))
col2<- gsub("FALSE","black",gsub("TRUE","red",color))
#
# デフォルト:par(mar=c(5,4,4,2)+0.1)
#par(fig=c(0,1,0,1),mar=c(3,5,3,2),family="serif")
# png("overlay01.png",width=800,height=800)
par(mfrow=c(1,1),fig=c(0,1,0,1),mar=c(3,12,3,2),family="serif")
b<- barplot(df[,"Tests /millionpeople"],horiz=T,col=col,xaxt="n",xlim=c(0,lim))
axis(side=1, at=axTicks(1), labels=formatC(axTicks(1), format="d", big.mark=','))
axis(2,at=b,labels=NA,col=col2,tck=-0.01)
text(x=par("usr")[1],y=b, labels = df[,1], col = col2,pos=2,xpd=T)
# bquote
title(bquote("Tests /million for COVID-19("~Positive>=.(min)~")"))
abline(v=jnum,lty=2,col="red")
arrows(x0=df[df[,1]=="Japan","Tests /millionpeople"]*4, y0=grep("TRUE",color)+0.5, x1=25000, y1=25,col="red", lty=1, lwd=5)
#par(mar=c(10,16,10,2.5),family="serif",new=T)
#右上
#par(fig=c(0.5,1,0.5,1),family="serif",new=T)
#右下
par(fig=c(0.5,1,0,0.5)+c(-0.1,-0.1,0.1,0.2),mar=c(3,6,4,2),family="serif",new=T)
b<- barplot(rev(df2[,3]),horiz=T,col=c(rep("pink",nrow(df2)-1),"red"),las=1)
axis(2,at=b,labels=NA,col=col2,tck=-0.01)
text(x=par("usr")[1],y=b, labels =rev(df2[,1]), col = c(rep("black",nrow(df2)-1),"red"),pos=2,xpd=T)
title("Tests /million:Prefectures of Japan\n(More than 20 COVID-19 cases reported)")
abline(v=jnum,lty=2,col="red")
# dev.off()