提交: 270df40bc2f57d6291010c6a0257bb13a0b6667c - myprojects-hg-reborn (hg)

Commit MetaInfo

• 修订版: 270df40bc2f57d6291010c6a0257bb13a0b6667c (tree)
• 时间: 2017-08-23 00:03:11
• 作者: Lorenzo Isella <lorenzo.isella@gmai...>
• Commiter: Lorenzo Isella

Log Message

I cleaned up the code and I added an aggregation by nace code. I also am using the cowplot library to save ggplot objects to pdf or png.

更改概述

• modified: R-codes/eurofund.R

差异

diff -r eec95040918c -r 270df40bc2f5 R-codes/eurofund.R
--- a/R-codes/eurofund.R	Tue Aug 22 15:36:48 2017 +0200
+++ b/R-codes/eurofund.R	Tue Aug 22 17:03:11 2017 +0200
@@ -6,6 +6,8 @@
 library(reshape2)
 library(magrittr)
 ## library(ggthemes)
+library(cowplot)
+
 library(haven)
 library(scales)
 library(ggthemes)
@@ -13,7 +15,6 @@
 library(ecp)
 library(ggthemes)
 
-
 options( scipen = 16 )
 
 source("/home/lorenzo/myprojects-hg/R-codes/stat_lib.R")
@@ -29,7 +30,7 @@
 
 
 
-df <- read_dta("eurofundata.dta") 
+df <- read_dta("eurofundata.dta")  %>% as.tibble %>% mutate(nace2digits= substrLeft(Nacecode, 2) )
 
 
 df_table_trade <- df  %>% filter(TypeofRestructuring %in%  trade_restructuring)  %>% group_by(Country, year) %>% summarise(total.count=n())
@@ -37,82 +38,22 @@
 
 df_table_all<- df   %>% group_by(Country, year) %>% summarise(total.count=n())
 
-## test <- df  %>%  filter(Country=="Cyprus") %>%  filter(TypeofRestructuring %in%  trade_restructuring)
-
-
-myformula <- paste("NumberEmployed", "~ year+  Country ") %>% as.formula()
-
-
-df_table1 <- df %>% filter(TypeofRestructuring %in%  trade_restructuring)  %>% {  aggregate(myformula, data=.  , FUN= sum, na.rm=T)}
-
-
-
-myformula <- paste("PlannedAVG   ", "~ year+  Country ") %>% as.formula()
-
-
-df_table2 <- df %>% filter(TypeofRestructuring %in%  trade_restructuring)  %>% {  aggregate(myformula, data=.  , FUN= sum, na.rm=T)}
-
-
-
-
 df_table2bis <- df %>% filter(TypeofRestructuring %in%  trade_restructuring)  %>% group_by(Country, year) %>% summarise(avg=sum(PlannedAVG, na.rm=T))
 
 
-
-
-df_table3 <- df %>% filter(TypeofRestructuring %in%  trade_restructuring)
-
-
-
-
-
-
-
-myformula <- paste("PlannedAVG   ", "~ year+  Country  ") %>% as.formula()
-
-
-df_table4 <- df %>% filter(TypeofRestructuring %in%  trade_restructuring)  %>% {  aggregate(myformula, data=.  , FUN= sum, na.rm=T)} %>% mutate(type="Trade") %>% filter(Country != "World")
-
-
-
-df_table5 <- df   %>% {  aggregate(myformula, data=.  , FUN= sum, na.rm=T)} %>% mutate(type="All") %>% filter(Country != "World")
-
-
-
-###  a better way to do this aggregation (and for several variables in one go)
-### is to use group_by and summarise. NB: I treat the missing data as zero.
-
+df_table3 <- df %>% filter(TypeofRestructuring %in%  trade_restructuring) %>% group_by(Country, year, nace2digits ) %>% summarise(nace2sum=sum(PlannedAVG, na.rm=T))  
 
 
 df_table4bis <- df %>% filter(TypeofRestructuring %in%  trade_restructuring) %>% group_by(Country, year) %>% summarise(avg=sum(PlannedAVG, na.rm=T),max=sum(PlannedJobReductionsmax, na.rm=T), min=sum(PlannedJobReductionsmin, na.rm=T), number_emp=sum(NumberEmployed, na.rm=T) ) %>% filter(Country != "World")%>% mutate(type="Trade")
 
-
-
-
-
-
-
 df_table5bis <- df %>%  group_by(Country, year) %>% summarise(avg=sum(PlannedAVG, na.rm=T),max=sum(PlannedJobReductionsmax, na.rm=T), min=sum(PlannedJobReductionsmin, na.rm=T), number_emp=sum(NumberEmployed, na.rm=T) ) %>% filter(Country != "World")%>% mutate(type="All")
 
 
-
 df_tot <- rbind(df_table4bis,df_table5bis)
 
 
 ## There is little difference between min, max and average, so there is no point in adding the min and max
 
-## myformula <- paste(" . ~ year+  Country ") %>% as.formula()
-
-
-## df_table6 <- df %>% filter(TypeofRestructuring %in%  trade_restructuring)  %>% select(c(PlannedAVG,PlannedJobReductionsmax, PlannedJobReductionsmin, year, Country ))  %>% {  aggregate(myformula, data=.  , FUN= sum, na.rm=F)} %>% filter(Country != "World")
-
-
-
-
-## df_table6bis <- df %>% filter(TypeofRestructuring %in%  trade_restructuring)  %>% select(c(PlannedAVG, year, Country ))  %>% {  aggregate(myformula, data=.  , FUN= sum, na.rm=T)} %>% filter(Country != "World")
-
-
-
 
 
 if (plot_loop == 1){
@@ -177,28 +118,7 @@
 ggsave("synoptic-trade2.pdf", gpl, width=15,height=15)
 
 
-
-
-
-
-
-## gpl <- ggplot(df_table3, aes(x = PlannedAVG, y = year, group = year)) +
-##   ## geom_joy(scale = 10, size = 0.25, rel_min_height = 0.03) +
-##   ##   theme_joy() +
-## geom_joy()+
     
-##   ## scale_x_continuous(limits=c(1, 200), expand = c(0.01, 0)) +
-##   ## scale_y_reverse(breaks=c(2000, 1980, 1960, 1940, 1920, 1900), expand = c(0.01, 0))
-## coord_cartesian(xlim=c(0,600))+
-
-##   scale_x_continuous() +
-##   scale_y_reverse()
-
-## ggsave("joyplot_average.pdf", gpl, width=10,height=5)
-
-
-
-
 lbls <- unique(df_tot$type)
 
 
@@ -276,10 +196,13 @@
 
 countries <- unique(df_tot$Country)
 
+## countries <- "Germany"
+    
 
 for (mycountry in countries){
 
-
+    print("mycountry is, ")
+    print(mycountry)
 
 temp <- df_tot %>% filter(Country == mycountry)
 
@@ -320,15 +243,87 @@
  ylab("Average Number Employees")
 
 
-fname <- paste(mycountry, "_single.pdf")
+fname <- paste(mycountry, "_single.pdf", sep="")
 
 ggsave(fname, gpl, width=10,height=5)
 
 
-fname <- paste(mycountry, "_single.png")
+fname <- paste(mycountry, "_single.png", sep="")
 
 ggsave(fname, gpl, width=10,height=5)
 
+
+
+temp2 <- df_table3 %>% filter(Country==mycountry)
+
+if (nrow(temp2)>0){
+    
+ gpl <- ggplot(temp2, aes(x=nace2digits, y=nace2sum,
+                        ## colour=indicator,
+                             ## fill=nace2digits
+                             ## group=country  
+                       )) +
+
+    
+    
+## theme(legend.position = 'right')+
+## geom_point(size=2, col="black") + 
+## geom_line(col="black")+
+
+## geom_point(size=2) + 
+
+geom_bar(position="dodge", stat="identity", alpha=1, fill="blue"
+         ) + 
+
+
+    
+#scale_y_continuous(limits=c(0.3,0.9),breaks=seq(0.3, 0.9, by=0.3))+
+#scale_y_continuous(breaks=pretty_breaks(n=5))+
+scale_y_continuous(breaks=pretty_breaks(n=5))+
+    
+## scale_x_continuous(breaks=integer_breaks(n=5))+
+    ## scale_x_continuous(breaks=data_g_temp$year
+    ##                )+
+
+my_ggplot_theme(c(0.13,0.8))+
+theme(legend.position = 'right')+
+
+facet_wrap( ~ year, ncol = 2, scales = "fixed"  )+
+
+## scale_fill_gdocs(NULL, breaks=lbls)+
+    
+labs(title=mycountry)+
+theme(plot.title = element_text(lineheight=.8, size=24, face="bold", vjust=1))+
+## theme(legend.text = element_text(vjust=1,lineheight=1 ))+
+## theme(legend.title = element_text(colour="black", size=16, face="bold"))+
+
+
+theme(legend.position = "none")+
+
+    
+
+    
+## scale_fill_manual("", breaks=lbls,  labels=c("Indirect Funding", "Funding from government"),values=col_seq) +
+
+## scale_color_manual("", breaks=lbls,   labels=c("Indirect Funding", "Funding from government"), values=col_seq) +
+
+
+
+
+xlab("NACE Division")+
+ ylab("Number of Employees")
+
+fname <- paste(mycountry, "_nace_division.pdf", sep="")
+
+    
+save_plot(fname, gpl,base_height=14,  base_aspect_ratio=2.5)
+
+    
+fname <- paste(mycountry, "_nace_division.png", sep="")
+
+save_plot(fname, gpl,base_height=14,  base_aspect_ratio=2.5)
+    
+}   
     
 }    
 }