Add consolidated historic data + crosswalked zips

code/consolidate1980Data.R

@@ -0,0 +1,200 @@
+
+# Author: Ashlynn Wimer
+# Date: 8/7/2023
+# About: This R script combines the 1980 historic data on 2010 geometries, and aggregates
+# from county level to state level.
+
+#### Libraries ----
+
+library(dplyr)
+library(stringr)
+
+#### Tracts ----
+
+ds01 <- read.csv('../data_final/DS01_T_1980_NHGIS.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=11, side='left', pad='0')) |>
+  select(-FIPS)
+
+# These two datasets were collected by Social Explorer. They feature data for
+# every 2010 census tract. However, the 1980 census did not collect that much
+# data -- they only collected 59187 tracts worth of data -- we discard the
+# excess data.
+
+ec03 <- read.csv('../data_final/EC03_T_1980_DC.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=11, side='left', pad='0')) |>
+  select(-FIPS, -X, -NAME, -year)
+hs01 <- read.csv('../data_final/HS01_T_1980_DC.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=11, side='left', pad='0')) |>
+  select(-FIPS, -X, -NAME, -year)
+
+# 1980 census tracts featured suppression of age data in small census tracts.
+# Thus, the totalPop variable has more people than the list of people whose
+# age was actually counted. 
+
+# The upshot of this is that calculating ChildrenP requires us to make a new
+# variable called "totalPop.age" which is just the number of people whose 
+# age was recorded.
+
+totalPop.age <- ds01$age0_4   + ds01$age5_14  + ds01$age15_44 + 
+                ds01$age45_54 + ds01$age55_59 + ds01$age60_64 + 
+                ds01$ageOv65
+
+ds01 <- ds01 |> 
+  mutate(childrenP = round(100 * (totalPop.age - age18_64 - ageOv65) / totalPop.age, 2))
+
+rm(totalPop.age)
+
+tracts <- merge(ds01, ec03, by='GEOID') |>
+  merge(hs01, by='GEOID') |>
+  select( 
+         GEOID,    totPop,   totUnits, age18_64,  age0_4, 
+         age5_14,  age15_19, age20_24, age15_44,  age45_54, 
+         age55_59, age60_64, ageOv65,  Age15_24P, Und45P, 
+         Ovr65P,   WhiteP,   BlackP,   HispP,     AmIndP,
+         AsianP,   PacIsP,   OtherP,   NoHSP,     childrenP, 
+         povP,     UnempP,   occP,     vacantP
+  )
+
+names <- read.csv('../data_raw/rename_tables/1980.csv')
+
+if (all(names(tracts) == names[,1])) {
+  colnames(tracts) <- names[,2]
+} else {
+  print("Names are misaligned!!")
+}
+
+rm(names)
+
+## Round off counts
+
+tracts <- tracts |> 
+  mutate(
+    TotPop   = round(TotPop),
+    TotUnits = round(TotUnits),
+    Age18_64 = round(Age18_64),
+    Age0_4   = round(Age0_4),
+    Age5_14  = round(Age5_14),
+    Age15_19 = round(Age15_19),
+    Age20_24 = round(Age20_24),
+    Age15_44 = round(Age15_44),
+    Age45_54 = round(Age45_54),
+    Age55_59 = round(Age55_59),
+    Age60_64 = round(Age60_64),
+    AgeOv65  = round(AgeOv65)
+  )
+
+write.csv(tracts, "../data_final/consolidated/T_1980.csv", row.names=F)
+
+#### Counties ---- 
+
+ds01 <- read.csv('../data_final/DS01_C_1980_NHGIS.csv') |>
+  mutate(GEOID = str_pad(GEOID, width=5, side='left', pad='0'),
+         childrenP = round(100 * (totPop - age18_64 - ageOv65) / totPop, 2) )
+
+ec03 <- read.csv('../data_final/EC03_C_1980_DC.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=5, side='left', pad='0')) |>
+  select(-FIPS, -NAME, -year)
+hs01 <- read.csv('../data_final/HS01_C_1980_DC.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=5, side='left', pad='0')) |>
+  select(-FIPS, -NAME, -X, -year)
+
+cnty <- merge(ds01, ec03, by='GEOID') |>
+  merge(hs01, by='GEOID') |>
+  mutate(
+    TotPop   = round(totPop),
+    TotUnits = round(totUnits),
+    Age18_64 = round(age18_64),
+    Age0_4   = round(age0_4),
+    Age5_14  = round(age5_14),
+    Age15_19 = round(age15_19),
+    Age20_24 = round(age20_24),
+    Age15_44 = round(age15_44),
+    Age45_54 = round(age45_54),
+    Age55_59 = round(age55_59),
+    Age60_64 = round(age60_64),
+    AgeOv65  = round(ageOv65)
+  )
+
+cntySave <- cnty |>
+  select( 
+    GEOID,    totPop,   totUnits, age18_64,  age0_4, 
+    age5_14,  age15_19, age20_24, age15_44,  age45_54, 
+    age55_59, age60_64, ageOv65,  Age15_24P, Und45P, 
+    Ovr65P,   WhiteP,   BlackP,   HispP,     AmIndP,
+    AsianP,   PacIsP,   OtherP,   NoHSP,     childrenP, 
+    povP,     UnempP,   occP,     vacantP
+  )
+
+names <- read.csv('../data_raw/rename_tables/1980.csv')
+
+if (all(names(cntySave) == names[,1])) {
+  colnames(cntySave) <- names[,2]
+} else {
+  print("Names are misaligned!!")
+}
+
+cntySave |> write.csv('C_1980.csv', row.names=F)
+
+##### States ----
+
+states <- cnty |> mutate(
+  STATEFP = substr(GEOID, start=1, stop=2)
+  ) |>
+  group_by(STATEFP) |>
+  summarize(
+    TotPop   = sum(totPop,             na.rm = T),
+    TotUnits = sum(totUnits,           na.rm = T),
+    Age18_64 = sum(age18_64,           na.rm = T),
+    Age0_4   = sum(age0_4,             na.rm = T),
+    Age5_14  = sum(age5_14,            na.rm = T),
+    Age15_19 = sum(age15_19,           na.rm = T),
+    Age20_24 = sum(age20_24,           na.rm = T),
+    Age15_44 = sum(age15_44,           na.rm = T),
+    Age45_54 = sum(age45_54,           na.rm = T),
+    Age55_59 = sum(age55_59,           na.rm = T),
+    Age60_64 = sum(age60_64,           na.rm = T),
+    AgeOv65  = sum(ageOv65,            na.rm = T),
+    WhitePop = sum(totPop * WhiteP,    na.rm = T),
+    BlackPop = sum(totPop * BlackP,    na.rm = T),
+    HispPop  = sum(totPop * HispP ,    na.rm = T),
+    AmIndPop = sum(totPop * AmIndP,    na.rm = T),
+    AsianPop = sum(totPop * AsianP,    na.rm = T),
+    PacIsPop = sum(totPop * PacIsP,    na.rm = T),
+    OtherPop = sum(totPop * OtherP,    na.rm = T),
+    NoHSPop  = sum(edSampl * NoHSP,    na.rm = T),
+    edSmpl   = sum(edSampl,            na.rm = T),
+    PovPop   = sum(povUni * povP,      na.rm = T),
+    povUni   = sum(povUni,             na.rm = T),
+    UnempPop = sum(labor * UnempP,     na.rm = T),
+    labor    = sum(labor,              na.rm = T),
+    OccTot   = sum(occP * totUnits,    na.rm = T),
+    VacTot   = sum(vacantP * totUnits, na.rm = T)
+  ) |>
+  mutate(
+    Age15_24P = round(100 * (Age15_19 + Age20_24) / TotPop, 2),
+    NoHsp     = round(NoHSPop  / edSmpl, 2),
+    WhiteP    = round(WhitePop / TotPop, 2),
+    BlackP    = round(BlackPop / TotPop, 2),
+    HispP     = round(HispPop  / TotPop, 2),
+    AmIndP    = round(AmIndPop / TotPop, 2),
+    AsianP    = round(AsianPop / TotPop, 2),
+    PacIsP    = round(PacIsPop / TotPop, 2),
+    OtherP    = round(OtherPop / TotPop, 2),
+    Und45P    = round(100 * (Age0_4 + Age5_14 + Age15_44) / (TotPop), 2),
+    Ovr65P    = round(100 * (AgeOv65) / TotPop, 2),
+    ChildrenP = round(100 * (TotPop - Age18_64 - AgeOv65) / TotPop, 2),
+    PovP      = round(PovPop / povUni, 2),
+    UnempP    = round(UnempPop / labor, 2),
+    OccP      = round(OccTot / TotUnits, 2),
+    VacantP   = round(VacTot / TotUnits, 2)
+  ) |>
+  select(
+    STATEFP,  TotPop,   TotUnits, Age18_64,  Age0_4, 
+    Age5_14,  Age15_19, Age20_24, Age15_44,  Age45_54, 
+    Age55_59, Age60_64, AgeOv65,  Age15_24P, Und45P, 
+    Ovr65P,   WhiteP,   BlackP,   HispP,     AmIndP,
+    AsianP,   PacIsP,   OtherP,   NoHsp,     ChildrenP, 
+    PovP,     UnempP,   OccP,     VacantP
+  )
+
+states |> write.csv("../data_final/consolidated/S_1980.csv", row.names = F)

code/consolidate1990Data.R

@@ -0,0 +1,187 @@
+
+# Author: Ashlynn Wimer
+# Date: 8/7/2023
+# About: This R script combines the 1990 historic data on 2010 geometries, and aggregates
+# from county level to state level.
+
+#### Libraries ----
+
+library(dplyr)
+library(stringr)
+
+
+#### Tracts ----
+
+# Read in data
+
+ds01 <- read.csv('../data_final/DS01_T_1990_NHGIS.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=11, side='left', pad='0')) |>
+  select(-FIPS)
+ec03 <- read.csv('../data_final/EC03_T_1990_DC.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=11, side='left', pad='0')) |>
+  select(-FIPS, -X, -NAME, -year)
+hs01 <- read.csv('../data_final/HS01_T_1990_DC.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=11, side='left', pad='0')) |>
+  select(-FIPS, -X, -NAME, -year)
+
+ds01 <- ds01 |> 
+  mutate(childrenP = round(100 * (totPop - age18_64 - agOver65) / totPop, 2))
+
+tracts <- merge(ds01, ec03, by='GEOID') |>
+  merge(hs01, by='GEOID') |>
+  select( 
+    GEOID,     totPop,   totUnits, age18_64,  age0_4, 
+    age5_14,   age15_19, age20_24, age15_44,  age45_49,
+    age50_54,  age55_59, age60_64, agOver65,   Age15_24P, 
+    Und45P,    Ovr65P,   WhiteP,   BlackP,    HispP,     
+    AmIndP,    AsianP,   PacIsP,   OtherP,    NoHSP,     
+    childrenP, povP,     UnempP,   occP,      vacantP
+  )
+
+names <- read.csv('../data_raw/rename_tables/1990.csv')
+
+if (all(names(tracts) == names[,1])) {
+  colnames(tracts) <- names[,2]
+} else {
+  print("Names are misaligned!!")
+}
+
+rm(names)
+
+## Round off counts
+
+tracts <- tracts |> 
+  mutate(
+    TotPop   = round(TotPop),
+    TotUnits = round(TotUnits),
+    Age18_64 = round(Age18_64),
+    Age0_4   = round(Age0_4),
+    Age5_14  = round(Age5_14),
+    Age15_19 = round(Age15_19),
+    Age20_24 = round(Age20_24),
+    Age15_44 = round(Age15_44),
+    Age45_49 = round(Age45_49),
+    Age50_54 = round(Age50_54),
+    Age55_59 = round(Age55_59),
+    Age60_64 = round(Age60_64),
+    AgeOv65  = round(AgeOv65)
+  )
+
+write.csv(tracts, "../data_final/consolidated/T_1990.csv", row.names=F)
+
+rm(tracts)
+
+#### Counties ---- 
+
+ds01 <- read.csv('../data_final/DS01_C_1990_NHGIS.csv') |>
+  mutate(GEOID = str_pad(GEOID, width=5, side='left', pad='0'),
+         childrenP = round(100 * (totPop - age18_64 - agOver65) / totPop, 2) )
+ec03 <- read.csv('../data_final/EC03_C_1990_DC.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=5, side='left', pad='0')) |>
+  select(-FIPS, -NAME, -year)
+hs01 <- read.csv('../data_final/HS01_C_1990_DC.csv') |>
+  mutate(GEOID = str_pad(FIPS, width=5, side='left', pad='0')) |>
+  select(-FIPS, -NAME, -X, -year)
+
+cnty <- merge(ds01, ec03, by='GEOID') |>
+  merge(hs01, by='GEOID') |>
+  mutate(
+    totPop   = round(totPop),
+    totUnits = round(totUnits),
+    age18_64 = round(age18_64),
+    age0_4   = round(age0_4),
+    age5_14  = round(age5_14),
+    age15_19 = round(age15_19),
+    age20_24 = round(age20_24),
+    age15_44 = round(age15_44),
+    age45_49 = round(age45_49),
+    age50_54 = round(age50_54),
+    age55_59 = round(age55_59),
+    age60_64 = round(age60_64),
+    agOver65  = round(agOver65)
+  )
+
+cntySave <- cnty |>
+  select( 
+    GEOID,     totPop,   totUnits, age18_64, age0_4, 
+    age5_14,   age15_19, age20_24, age15_44, age45_49,
+    age50_54,  age55_59, age60_64, agOver65, Age15_24P, 
+    Und45P,    Ovr65P,   WhiteP,   BlackP,   HispP,
+    AmIndP,    AsianP,   PacIsP,   OtherP,   NoHSP,
+    childrenP, povP,     UnempP,   occP,     vacantP
+  )
+
+names <- read.csv('../data_raw/rename_tables/1990.csv')
+
+if (all(names(cntySave) == names[,1])) {
+  colnames(cntySave) <- names[,2]
+} else {
+  print("Names are misaligned!!")
+}
+
+cntySave |> write.csv('../data_final/consolidated/C_1990.csv', row.names=F)
+
+##### States ----
+
+states <- cnty |> mutate(
+  STATEFP = substr(GEOID, start=1, stop=2)
+) |>
+  group_by(STATEFP) |>
+  summarize(
+    TotPop   = sum(totPop,             na.rm = T),
+    TotUnits = sum(totUnits,           na.rm = T),
+    Age18_64 = sum(age18_64,           na.rm = T),
+    Age0_4   = sum(age0_4,             na.rm = T),
+    Age5_14  = sum(age5_14,            na.rm = T),
+    Age15_19 = sum(age15_19,           na.rm = T),
+    Age20_24 = sum(age20_24,           na.rm = T),
+    Age15_44 = sum(age15_44,           na.rm = T),
+    Age45_49 = sum(age45_49,           na.rm = T),
+    Age50_54 = sum(age50_54,           na.rm = T),
+    Age55_59 = sum(age55_59,           na.rm = T),
+    Age60_64 = sum(age60_64,           na.rm = T),
+    AgeOv65  = sum(agOver65,            na.rm = T),
+    WhitePop = sum(totPop * WhiteP,    na.rm = T),
+    BlackPop = sum(totPop * BlackP,    na.rm = T),
+    HispPop  = sum(totPop * HispP ,    na.rm = T),
+    AmIndPop = sum(totPop * AmIndP,    na.rm = T),
+    AsianPop = sum(totPop * AsianP,    na.rm = T),
+    PacIsPop = sum(totPop * PacIsP,    na.rm = T),
+    OtherPop = sum(totPop * OtherP,    na.rm = T),
+    NoHSPop  = sum(edSampl * NoHSP,    na.rm = T),
+    edSmpl   = sum(edSampl,            na.rm = T),
+    PovPop   = sum(povUni * povP,      na.rm = T),
+    povUni   = sum(povUni,             na.rm = T),
+    UnempPop = sum(labor * UnempP,     na.rm = T),
+    labor    = sum(labor,              na.rm = T),
+    OccTot   = sum(occP * totUnits,    na.rm = T),
+    VacTot   = sum(vacantP * totUnits, na.rm = T)
+  ) |>
+  mutate(
+    Age15_24P = round(100 * (Age15_19 + Age20_24) / TotPop, 2),
+    NoHsp     = round(NoHSPop  / edSmpl, 2),
+    WhiteP    = round(WhitePop / TotPop, 2),
+    BlackP    = round(BlackPop / TotPop, 2),
+    HispP     = round(HispPop  / TotPop, 2),
+    AmIndP    = round(AmIndPop / TotPop, 2),
+    AsianP    = round(AsianPop / TotPop, 2),
+    PacIsP    = round(PacIsPop / TotPop, 2),
+    OtherP    = round(OtherPop / TotPop, 2),
+    Und45P    = round(100 * (Age0_4 + Age5_14 + Age15_44) / (TotPop), 2),
+    Ovr65P    = round(100 * (AgeOv65) / TotPop, 2),
+    ChildrenP = round(100 * (TotPop - Age18_64 - AgeOv65) / TotPop, 2),
+    PovP      = round(PovPop / povUni, 2),
+    UnempP    = round(UnempPop / labor, 2),
+    OccP      = round(OccTot / TotUnits, 2),
+    VacantP   = round(VacTot / TotUnits, 2)
+  ) |>
+  select(
+    STATEFP,   TotPop,   TotUnits, Age18_64,  Age0_4, 
+    Age5_14,   Age15_19, Age20_24, Age15_44,  Age45_49,
+    Age50_54,  Age55_59, Age60_64, AgeOv65,  Age15_24P, 
+    Und45P,    Ovr65P,   WhiteP,   BlackP,   HispP,     
+    AmIndP,    AsianP,   PacIsP,   OtherP,   NoHsp,     
+    ChildrenP, PovP,     UnempP,   OccP,     VacantP
+  )
+
+states |> write.csv("../data_final/consolidated/S_1990.csv", row.names=F)