git pull error message asking to commit changes first

Merge branch 'blais_replication' of https://github.com/sjkiss/CES_Analysis into blais_replication # Conflicts: # R_Scripts/1_master_file.R # R_Scripts/2_11_ces00_recode.R
sjkiss · Jun 17, 2020 · 7b1aee3 · 7b1aee3
2 parents ba11ccd + b8b0907
commit 7b1aee3
Show file tree

Hide file tree

Showing 53 changed files with 533 additions and 876 deletions.
diff --git a/Plots/IM1_union_both&sector_ndp_coefficients.png b/Plots/IM1_union_both&sector_ndp_coefficients.png
diff --git a/Plots/IM2_union_both&sector_ndp_coefficients.png b/Plots/IM2_union_both&sector_ndp_coefficients.png
diff --git a/Plots/IM3_degree&income_ndp_coefficients.png b/Plots/IM3_degree&income_ndp_coefficients.png
diff --git a/Plots/M1_ndp_by_sector.png b/Plots/M1_ndp_by_sector.png
diff --git a/Plots/M1_union_both&ndp_coefficients.png b/Plots/M1_union_both&ndp_coefficients.png
diff --git a/Plots/M2_union_both&ndp_coefficients.png b/Plots/M2_union_both&ndp_coefficients.png
diff --git a/Plots/M3_union_both&ndp_coefficients.png b/Plots/M3_union_both&ndp_coefficients.png
diff --git a/Plots/M4_union_both&ndp_coefficients.png b/Plots/M4_union_both&ndp_coefficients.png
diff --git a/Plots/M5_union_both&ndp_coefficients.png b/Plots/M5_union_both&ndp_coefficients.png
diff --git a/Plots/M6_union_both&ndp_coefficients.png b/Plots/M6_union_both&ndp_coefficients.png
diff --git a/Plots/M7_union_both&ndp_coefficients.png b/Plots/M7_union_both&ndp_coefficients.png
diff --git a/Plots/M8_union_both&ndp_coefficients.png b/Plots/M8_union_both&ndp_coefficients.png
diff --git a/Plots/degree_ndp_coefficients.png b/Plots/degree_ndp_coefficients.png
diff --git a/Plots/logit_union_ndp.png b/Plots/logit_union_ndp.png
diff --git a/Plots/percent_ndp_degree.png b/Plots/percent_ndp_degree.png
diff --git a/Plots/percent_ndp_union.png b/Plots/percent_ndp_union.png
diff --git a/Plots/piketty_party_vote_by_degree.png b/Plots/piketty_party_vote_by_degree.png
diff --git a/Plots/sector_conservative_1968_2015.png b/Plots/sector_conservative_1968_2015.png
diff --git a/Plots/sector_liberal_1968_2015.png b/Plots/sector_liberal_1968_2015.png
diff --git a/Plots/sector_ndp_1968_2015.png b/Plots/sector_ndp_1968_2015.png
diff --git a/Plots/share_degree_ces.png b/Plots/share_degree_ces.png
diff --git a/Plots/union_both&ndp_coefficients.png b/Plots/union_both&ndp_coefficients.png
diff --git a/Plots/union_ndp_coefficients.png b/Plots/union_ndp_coefficients.png
diff --git a/R_Scripts/2_0_recode_file.R b/R_Scripts/2_0_recode_file.R
@@ -17,13 +17,18 @@ nrow(ces7980)==2761
 nrow(ces84)==3377
 nrow(ces88)==3609
 ####Check that the datasets do not contain the recoded variable names
+tail(names(ces74))
 tail(names(ces84))
 tail(names(ces0411))
 tail(names(ces15phone))
-
+#extra checks for ces74
+table(ces74$sector)
+table(ces74$union)
 ## Note: If it appears that the loaded datasets include recoded variable names at this point you may  neee do uncomment the next line, and run it once
 #rm(list = ls(all.names = TRUE)) #will clear all objects includes hidden objects.
 #### and then go back up and re-run the code, not re-running the rm() line above again. 
+
+
 #load the here() library
 library(here)
 ###How to use the here()package

diff --git a/R_Scripts/2_12_ces0411_recode.R b/R_Scripts/2_12_ces0411_recode.R
@@ -2,65 +2,72 @@
 library(tidyverse)
 library(car)
 library(labelled)
-library(cesdata)
+
 #load data
 data("ces0411")
 
-#------------------------------------------------------------------------------
-# Gender is the same variable for all elections 2004-11
 
-#recode Gender (GENDER)
+####recode Gender (GENDER)####
+#Gender is the same variable for all elections 2004-11
+
 look_for(ces0411, "sex")
 ces0411$male<-Recode(ces0411$GENDER, "1=1; 5=0")
 val_labels(ces0411$male)<-c(Female=0, Male=1)
 #checks
 val_labels(ces0411$male)
 table(ces0411$male)
 
-#------------------------------------------------------------------------------
-###Recode 2004 1st
-
+##### Union ####
 #recode Union Respondent (ces04_CPS_S6A)
-ces0411$union04<-Recode(ces0411$ces04_CPS_S6A, "1=1; 5=0; else=NA")
+ces0411$union04<-Recode(ces0411$ces04_CPS_S6A, 
+                        "1=1; 5=0; else=NA")
 val_labels(ces0411$union04)<-c(None=0, Union=1)
 #checks
 val_labels(ces0411$union04)
 table(ces0411$union04)
 
 #recode Union Combined (ces04_CPS_S6A and ces04_CPS_S6B)
-ces0411 %>% 
+table(ces0411$ces04_CPS_S6A, ces0411$ces04_CPS_S6B, useNA = "ifany")
+802+259+982
+# Note! The household union question was only asked if the respondent was not a member of a union or if they didn't answer!!
+ces0411 %>%
   mutate(union_both04=case_when(
     #If the person is in a union OR if the household is in a union, then they get a 1
-    ces04_CPS_S6A==1 | ces04_CPS_S6B==1 ~ 1,
-    #If the person is not in a union AND if the household is not in a union then thety get a 0 
-    ces04_CPS_S6A==5 | ces04_CPS_S6B==5 ~ 0,
+    ces04_CPS_S6A==1 |    ces04_CPS_S6B==1 ~ 1,
+    #If the person is not in a union AND if the household is not in a union then theyy get a 0
+     ces04_CPS_S6B==5 ~ 0,
     ces04_CPS_S6A==8 & ces04_CPS_S6B==8 ~ NA_real_,
     ces04_CPS_S6A==9 & ces04_CPS_S6B==9 ~ NA_real_,
+    TRUE ~0
   ))->ces0411
 
+table(ces0411$union_both04, useNA = "ifany")
+1061+984
 val_labels(ces0411$union_both04)<-c(None=0, Union=1)
 
 #checks
 val_labels(ces0411$union_both04)<-c(None=0, Union=1)
 table(ces0411$union_both04)
-table( as_factor(ces0411$ces04_CPS_S6A), as_factor(ces0411$union04))
-table( as_factor(ces0411$ces04_CPS_S6B), as_factor(ces0411$union04))
-table( as_factor(ces0411$ces04_CPS_S6A), as_factor(ces0411$union_both04))
-table( as_factor(ces0411$ces04_CPS_S6B), as_factor(ces0411$union_both04))
 
-table(as_factor(ces0411$union04), as_factor(ces0411$ces04_CPS_S6B))
-table(as_factor(ces0411$union_both04), as_factor(ces0411$ces04_CPS_S6A))
-table(as_factor(ces0411$union_both04), as_factor(ces0411$ces04_CPS_S6B))
+ces0411 %>% 
+  select(ces04_CPS_S6A, ces04_CPS_S6B, union_both04) %>% 
+  group_by(ces04_CPS_S6A, ces04_CPS_S6B, union_both04) %>% 
+  summarize(n=n())
+
+# Some checks
+table( as_factor(ces0411$ces04_CPS_S6A), as_factor(ces0411$ces04_CPS_S6B), useNA = "ifany")
+table(as_factor(ces0411$union_both04), as_factor(ces0411$ces04_CPS_S6A), useNA = "ifany")
+table(as_factor(ces0411$union_both04), as_factor(ces0411$ces04_CPS_S6B), useNA = "ifany")
 
-#recode Education (ces04_CPS_S3)
+####Education (ces04_CPS_S3)####
 look_for(ces0411, "education")
 ces0411$degree04<-Recode(ces0411$ces04_CPS_S3, "9:11=1; 1:8=0; else=NA")
 val_labels(ces0411$degree04)<-c(nodegree=0, degree=1)
 #checks
 val_labels(ces0411$degree04)
 table(ces0411$degree04)
 
-#recode Region (ces04_PROVINCE)
+#####Region (ces04_PROVINCE)####
 look_for(ces0411, "province")
 ces0411$region04<-Recode(ces0411$ces04_PROVINCE, "10:13=1; 35=2; 46:59=3; 4=NA; else=NA")
 val_labels(ces0411$region04)<-c(Atlantic=1, Ontario=2, West=3)
@@ -209,16 +216,17 @@ ces0411 %>%
     #If the person is in a union OR if the household is in a union, then they get a 1
     ces06_CPS_S6A==1 | ces06_CPS_S6B==1 ~ 1,
     #If the person is not in a union AND if the household is in a union, then they get a 0
-    ces06_CPS_S6A==5 | ces06_CPS_S6B==5 ~ 0,
+ces06_CPS_S6B==5 ~ 0,
     ces06_CPS_S6A==8 & ces06_CPS_S6B==8 ~ NA_real_,
     ces06_CPS_S6A==9 & ces06_CPS_S6B==9 ~ NA_real_,
+        TRUE ~0
   ))->ces0411
 
 val_labels(ces0411$union_both06)<-c(None=0, Union=1)
 #checks
 val_labels(ces0411$union_both06)
 table(ces0411$union_both06)
-
+table(as_factor(ces0411$ces06_CPS_S6A), as_factor(ces0411$union_both06), useNA = "ifany")
 #recode Education (ces06_CPS_S3)
 look_for(ces0411, "education")
 ces0411$degree06<-Recode(ces0411$ces06_CPS_S3, "9:11=1; 1:8=0; else=NA")
@@ -272,11 +280,14 @@ table(ces0411$language06)
 
 #recode Non-charter Language (ces06_CPS_S17)
 look_for(ces0411, "language")
+table(ces0411$ces06_CPS_S17)
 ces0411$non_charter_language06<-Recode(ces0411$ces06_CPS_S17, "1:5=0; 8:64=1; 65:66=0; 95:97=1; else=NA")
 val_labels(ces0411$non_charter_language06)<-c(Charter=0, Non_Charter=1)
 #checks
 val_labels(ces0411$non_charter_language06)
-table(ces0411$non_charter_language06)
+table(ces0411$ces06_CPS_S17, ces0411$non_charter_language06, useNA = "ifany")
+
+table(ces0411$survey, ces0411$non_charter_language06)
 
 #recode Employment (ces06_CPS_S4)
 look_for(ces0411, "employed")
@@ -324,6 +335,7 @@ table(ces0411$party_id06)
 
 #recode Vote (ces06_PES_B4A and ces06_PES_B4B) 
 look_for(ces0411, "party did you vote")
+ces0411$ces06_PES_B4A
 ces0411 %>% 
   mutate(vote06=case_when(
     ces06_PES_B4A==1 | ces06_PES_B4B==1 ~ 1,
@@ -333,7 +345,7 @@ ces0411 %>%
     ces06_PES_B4A==0 | ces06_PES_B4B==0 ~ 0,
     ces06_PES_B4A==4 | ces06_PES_B4B==4 ~ 4,
   ))->ces0411
-
+table(ces0411$ces06_PES_B4A, ces0411$vote06)
 val_labels(ces0411$vote06)<-c(Other=0, Liberal=1, Conservative=2, NDP=3, Bloc=4, Green=5)
 #checks
 val_labels(ces0411$vote06)
@@ -375,16 +387,17 @@ ces0411 %>%
     #If the person is in a union OR if the household is in a union, then they get a 1
     ces08_CPS_S6A==1 | ces08_CPS_S6B==1 ~ 1,
     #If the person is in a union AND if the household is in a union, then they get a 1
-    ces08_CPS_S6A==5 | ces08_CPS_S6B==5 ~ 0,
+    ces08_CPS_S6B==5 ~ 0,
     ces08_CPS_S6A==8 & ces08_CPS_S6B==8 ~ NA_real_,
     ces08_CPS_S6A==9 & ces08_CPS_S6B==9 ~ NA_real_,
+    TRUE ~ 0
   ))->ces0411
 
 val_labels(ces0411$union_both08)<-c(None=0, Union=1)
 #checks
 val_labels(ces0411$union_both08)
 table(ces0411$union_both08)
-
+table(as_factor(ces0411$ces08_CPS_S6A), as_factor(ces0411$union_both08), useNA = "ifany")
 #recode Education (ces08_CPS_S3)
 look_for(ces0411, "education")
 ces0411$degree08<-Recode(ces0411$ces08_CPS_S3, "9:11=1; 1:8=0; else=NA")
@@ -444,6 +457,7 @@ val_labels(ces0411$non_charter_language08)<-c(Charter=0, Non_Charter=1)
 #checks
 val_labels(ces0411$non_charter_language08)
 table(ces0411$non_charter_language08)
+table(ces0411$survey, ces0411$non_charter_language08)
 
 #recode Employment (ces08_CPS_S4)
 look_for(ces0411, "employed")
@@ -536,9 +550,10 @@ ces0411 %>%
   mutate(union_both11=case_when(
     #If the person is in a union OR if the household is in a union, then they get a 1
     PES11_93==1 | PES11_94==1 ~ 1,
-    PES11_93==5 | PES11_94==5 ~ 0,
+     PES11_94==5 ~ 0,
     PES11_93==8 & PES11_94==8 ~ NA_real_,
     PES11_93==9 & PES11_94==9 ~ NA_real_,
+    TRUE~0
   ))->ces0411
 
 table(as_factor(ces0411$union_both11), as_factor(ces0411$PES11_93))
@@ -674,3 +689,4 @@ val_labels(ces0411$income11)<-c(Lowest=1, Lower_Middle=2, MIddle=3, Upper_Middle
 #checks
 val_labels(ces0411$income11)
 table(ces0411$income11)
+
diff --git a/R_Scripts/2_13_ces15_recode.R b/R_Scripts/2_13_ces15_recode.R
@@ -22,15 +22,21 @@ table(ces15phone$union)
 ces15phone %>% 
   mutate(union_both=case_when(
     PES15_93==1 | PES15_94==1 ~ 1,
-    PES15_93==5 | PES15_94==5 ~ 0,
+    PES15_93==5 & PES15_94==5 ~ 0,
     PES15_93==8 & PES15_94==8 ~ NA_real_,
     PES15_93==9 & PES15_94==9 ~ NA_real_,
+    TRUE ~ 0,
   ))->ces15phone
 
 val_labels(ces15phone$union_both)<-c(None=0, Union=1)
 #checks
 val_labels(ces15phone$union_both)
+table(ces15phone$union_both, useNA="ifany")
+table(ces15phone$PES15_93, ces15phone$union_both, useNA="ifany")
+table(ces15phone$PES15_93, ces15phone$PES15_94, useNA="ifany")
 table(ces15phone$union_both)
+table(ces15phone$PES15_93, ces15phone$union_both, useNA="ifany")
+table(ces15phone$PES15_94, ces15phone$union_both, useNA="ifany")
 
 #recode Education (CPS15_79)
 look_for(ces15phone, "education")
@@ -82,7 +88,7 @@ table(ces15phone$language)
 #recode Non-charter Language (CPS15_90)
 look_for(ces15phone, "language")
 ces15phone$non_charter_language<-Recode(ces15phone$CPS15_90, "1:5=0; 8:64=1; 65=0; 95:97=1; else=NA")
-val_labels(ces15phone$non_charter_language15)<-c(Charter=0, Non_Charter=1)
+val_labels(ces15phone$non_charter_language)<-c(Charter=0, Non_Charter=1)
 #checks
 val_labels(ces15phone$non_charter_language)
 table(ces15phone$non_charter_language)

diff --git a/R_Scripts/2_2_ces68_recode.R b/R_Scripts/2_2_ces68_recode.R
@@ -29,19 +29,21 @@ ces68 %>%
     var363==3 | var379==3 ~ 1,
     var363==4 | var379==4 ~ 1,
     #This should only be missing if BOTH are not members, right?
-    #Note var379 is the spousal activity variable, by setting it to be less than zero, I am including people who anaswered 1 (not a member) and 0 the mysterious unknown category
     var363==1 | var379==1 ~ 0,
+        #Note var379 is the spousal activity variable
+        var379==0 ~ 0,
     #This should only be missing if BOTH are no reply, right?
-    var363==5 | var379==5 ~ NA_real_,
-    var379==0 ~ 0,
+    var363==5 | var379==5 ~ NA_real_
+
   ))->ces68
 
 val_labels(ces68$union_both)<-c(None=0, Union=1)
 #checks
 val_labels(ces68$union_both)
 table(ces68$union_both)
 table(ces68$union_both, ces68$var363)
-table(ces68$union_both, ces68$var379)
+table(ces68$union, ces68$var363,useNA = "ifany")
+table(ces68$union_both, ces68$var379,useNA = "ifany")
 
 #recode Education (var334)
 ces68$degree<-Recode(ces68$var334, "17:20=1; 25:26=1; 1:16=0; 21:24=0; 27=0; 30=NA")

diff --git a/R_Scripts/2_4_ces74_recode.R b/R_Scripts/2_4_ces74_recode.R
@@ -1,8 +1,9 @@
 #File to Recode 1974 CES Data for 1974 election 
-library(cesdata)
+
 #load data
 data("ces74")
 nrow(ces74)
+
 #recode Gender (V480)
 look_for(ces74, "sex")
 ces74$male<-Recode(ces74$V480, "1=1; 2=0; 9=NA")
@@ -13,21 +14,25 @@ table(ces74$male)
 
 #recode Union Household (V477)
 look_for(ces74, "union")
-
-ces74$union<-Recode(ces74$V476, "1=1; 2=0; 8=NA")
+ces74$union<-Recode(ces74$V477, "1=1; 2=0; 8=0")
 val_labels(ces74$union)<-c(None=0, Union=1)
 #checks
 val_labels(ces74$union)
 table(ces74$union)
 table(ces74$V476,ces74$V477)
 ces74$V477
 ces74$V478
+
 #Union Combined variable (identical copy of union)
 ces74$union_both<-ces74$union
+#checks
+val_labels(ces74$union_both)
+table(ces74$union_both)
+
 #recode Education (V414)
 look_for(ces74, "school")
 look_for(ces74, "degree")
-ces74$degree<-Recode(ces74$V417, "25=1; 0:13=0")
+ces74$degree<-Recode(ces74$V414, "25=1; 0:13=0")
 val_labels(ces74$degree)<-c(nodegree=0, degree=1)
 #checks
 val_labels(ces74$degree)
@@ -92,8 +97,16 @@ table(ces74$employment)
 #recode Sector (V386)
 look_for(ces74, "sector")
 look_for(ces74, "business")
+ces74 %>% 
+  mutate(sector=case_when(
+    V395==69.25 ~ 1,
+    V395==71.77 ~ 1,
+    V386==13 ~ 1,
+    V386> 0 & V386 < 13 ~ 0,
+    V381> 0 & V381 < 7 ~ 0,
+    V381==50 ~ 0,
+  ))->ces74
 
-ces74$sector<-Recode(ces74$V386, "13=1; 1:12=0; else=NA")
 val_labels(ces74$sector)<-c(Private=0, Public=1)
 #checks
 val_labels(ces74$sector)
@@ -130,7 +143,6 @@ val_labels(ces74$income)<-c(Lowest=1, Lower_Middle=2, MIddle=3, Upper_Middle=4,
 #checks
 val_labels(ces74$income)
 table(ces74$income)
-names(ces74)
 
 #recode Community Size (V9)
 look_for(ces74, "community")
@@ -141,3 +153,5 @@ val_labels(ces74$size)<-c(Rural=1, Under_10K=2, Under_100K=3, Under_500K=4, City
 #checks
 val_labels(ces74$size)
 table(ces74$size)
+
+names(ces74)
diff --git a/R_Scripts/2_5_ces7980_recode.R b/R_Scripts/2_5_ces7980_recode.R
@@ -112,6 +112,10 @@ ces7980$V1471
 table(ces7980$V1471)
 ces7980 %>% 
   mutate(sector=case_when(
+    #teachers and nurses are added to public sector
+    V1484==6925 ~ 1,
+    V1484==7177 ~ 1,
+    V1484==7230 ~ 1,
     #all government employees go to public sector
     V1473==13 ~ 1,
     #all non-government employees go to zero
@@ -161,7 +165,7 @@ val_labels(ces7980$income)<-c(Lowest=1, Lower_Middle=2, Middle=3, Upper_Middle=4
 val_labels(ces7980$income)
 table(ces7980$income)
 
-
+#--------------------------------------------------------------------------------------------------------------------
 ####1980
 #recode Gender (V2156)
 look_for(ces7980, "sex")
@@ -232,6 +236,7 @@ val_labels(ces7980$party_id80)<-c(Other=0, Liberal=1, Conservative=2, NDP=3)
 val_labels(ces7980$party_id80)
 table(ces7980$party_id80)
 table(ces7980$party_id, ces7980$party_id80)
+
 #recode Vote (V2062)
 look_for(ces7980, "vote")
 ces7980$vote80<-Recode(ces7980$V2062, "1=1; 2=2; 3=3; 4:5=0; else=NA")

diff --git a/R_Scripts/2_7_ces84_recode.R b/R_Scripts/2_7_ces84_recode.R
@@ -104,17 +104,20 @@ look_for(ces84, "employment")
 
 ces84 %>% 
   mutate(sector=case_when(
+    #what does your company do (government)
     VAR530==13 ~1,
+    # VAR530<13~0,
+    #     VAR524 >1 ~ 0,
+    #assume these are teachers and nurses
       VAR526> 2710 & VAR526 < 2800 ~ 1,
     VAR526> 3129 & VAR526 < 3136 ~ 1,
-    VAR524 >1 ~ 0,
-    VAR530==99 ~NA_real_ ,
+ #   VAR530==99 ~NA_real_ ,
+ #all else gets as per Blais' footnote (reading between the lines)
     TRUE ~ 0
   ))->ces84
-ces84$VAR524
-ces84$VAR530
-table(ces84$VAR524, ces84$sector, useNA='ifany')
-table(ces84$VAR524, ces84$VAR530, useNA="ifany")
+
+table(as_factor(ces84$VAR524), ces84$sector, useNA='ifany')
+table(as_factor(ces84$VAR524), as_factor(ces84$VAR530), useNA="ifany")
 val_labels(ces84$sector)<-c(Private=0, Public=1)
 ces84$sector
 #checks