Added final? IATRB models

vibe-lab-gsd · Jun 27, 2024 · c83d9a2 · c83d9a2
1 parent 5f0d0d2
commit c83d9a2
Show file tree

Hide file tree

Showing 58 changed files with 11,499 additions and 24 deletions.
diff --git a/compare_nests.png b/compare_nests.png
diff --git a/data-assembly/survey-quintile-example.Rmd b/data-assembly/survey-quintile-example.Rmd
@@ -55,18 +55,26 @@ hh_2017 <- here("nhts",
   # filter rows with missing value for income
   filter(income_k > 0) |> 
   # create survey object with weights
-  as_survey_design(weights = WTHHFIN) |> 
-  # Calculate quintile thresholds
-  mutate(quint_break = survey_quantile(income_k, 
-                                       quantiles = c(0.2, 
-                                                     0.4, 
-                                                     0.6, 
-                                                     0.8))) |>
+  as_survey_design(weights = WTHHFIN)
+```
+
+
+```{r}
+threshholds <- survey::svyquantile(~income_k,
+                                   design = hh_2017,
+                                   quantiles = c(0.2, 0.4, 0.6, 0.8))
+
+```
+
+
+```{r}
+
+hh_2017 <- hh_2017 |>
   # compare income_k value to thresholds to get income quintile of each HH
-  mutate(inc_quint = case_when(income_k < quint_break$`_q20`[1] ~ "q1",
-                               income_k < quint_break$`_q40`[1] ~ "q2",
-                               income_k < quint_break$`_q60`[1] ~ "q3",
-                               income_k < quint_break$`_q80`[1] ~ "q4",
+  mutate(inc_quint = case_when(income_k < threshholds$income_k[1] ~ "q1",
+                               income_k < threshholds$income_k[2] ~ "q2",
+                               income_k < threshholds$income_k[3] ~ "q3",
+                               income_k < threshholds$income_k[4] ~ "q4",
                                TRUE ~ "q5"))
 ```
 

diff --git a/data-assembly/upper-income-bin.R b/data-assembly/upper-income-bin.R
@@ -0,0 +1,33 @@
+library(tidyverse)
+library(here) 
+library(srvyr)
+library(survey)
+
+ipums_data <- here("data",
+                   "usa_00009.csv.gz") |>
+  read_csv() |>
+  filter(GQ != 3 & GQ != 4) 
+
+data_2009 <- ipums_data |>
+  filter(YEAR == 2009 & HHINCOME > 100000) |>
+  group_by(SERIAL) |>
+  summarise(CLUSTER = first(CLUSTER),
+            STRATA = first(STRATA),
+            HHWT = first(HHWT),
+            HHINCOME = first(HHINCOME)) |>
+  as_survey_design(ids = CLUSTER,
+                   weights = HHWT)
+
+survey::svyquantile(~HHINCOME, data_2009, 0.5)
+
+data_2017 <- ipums_data |>
+  filter(YEAR == 2017 & HHINCOME > 200000) |>
+  group_by(SERIAL) |>
+  summarise(CLUSTER = first(CLUSTER),
+            STRATA = first(STRATA),
+            HHWT = first(HHWT),
+            HHINCOME = first(HHINCOME)) |>
+  as_survey_design(ids = CLUSTER,
+                   weights = HHWT)
+
+survey::svyquantile(~HHINCOME, data_2017, 0.5)