svi.qmd

---
execute:
  warning: false
  error: false
  cache: true
---

# Social Vulnerability Index

## Data Procurement
*   Raw 2016-2012 ACS data used for calculations were downloaded from the `findSVI` package using the `tidycensus` dependency.  
*   ZCTA to County Crosswalk was downloaded from [https://www2.census.gov/geo/docs/maps-data/data/rel/zcta_county_rel_10.txt](https://www2.census.gov/geo/docs/maps-data/data/rel/zcta_county_rel_10.txt)

## Data Processing 
The `findSVI` package calculates SVI percentiles for a geography of choice. In this case, we used ZCTAs.

1.  The `find_svi` command was used to both download 2012-2016 ACS data at the ZCTA level and calculate the SVI quartiles as according to the CDC/ASTDR SVI documentation.
2.  Utilizing the ZCTA to county crosswalk, which contains a list of every ZCTA and its intersecting states, the `find_svi` results were limited to those ZCTAs which intersect with Texas and Louisiana
3.  Results were output to a parquet

More information about the `findSVI` package can be found [here](https://heli-xu.github.io/findSVI/).

## Data Output
-   `TXLA_ZCTA_SVI1216.parquet`

## Code
The variables used in this calculation are as follows:
```{r}
#| output: asis
library(findSVI)
library(tidycensus)
library(tidyverse)

acs2016vars <- c("acs5", "acs5/subject", "acs5/profile", "acs5/cprofile") %>%
  map_df(~ load_variables(2016, .x) %>% mutate(table = .x)) %>%
  select(-concept, -geography)

census_variables_2016 %>% 
  map(~ filter(acs2016vars, name %in% str_sub(.x, end = -2))) %>%
  map(knitr::kable)
```

``` {r}
#| include: false
#| eval: false

# this code generates the same data as get_census_data
df <- census_variables_2016 %>% 
  map(~ filter(acs2019vars, name %in% str_sub(.x, end = -2))) %>%
  list_rbind() %>%
  distinct() %>%
  group_by(table) %>%
  group_split() %>%
  map(~ get_acs("zcta", variables = .x$name, year = 2019, output = "wide")) %>%
  list_cbind() %>%
  rename(GEOID = GEOID...1, NAME = NAME...2) %>%
  select(-matches("\\d+$"))
```

This single command downloads the 2015-2019 ACS5 at the zcta level and calculates the SVI theme percentiles for the entire country
```{r}
svi <- find_svi(year = 2016, geography = "zcta", full.table = TRUE)
```

Calculated SVI is filtered to the state of Texas (FIPS: 48) and Louisiana (FIPS: 22)
```{r}
#| output: false

library(sf)
library(tigris)

ZCTAs <- read_csv("https://www2.census.gov/geo/docs/maps-data/data/rel/zcta_county_rel_10.txt") %>%
  filter(STATE %in% c(48, 22)) %>%
  distinct(ZCTA5) %>%
  rename(GEOID = ZCTA5)

ZCTAs_sf <- ZCTAs %>%
  left_join(zctas(cb = FALSE, year = 2017), by = join_by(GEOID == GEOID10)) %>%
  st_as_sf() %>%
  select(-c(ZCTA5CE10, CLASSFP10, MTFCC10, FUNCSTAT10, ALAND10, AWATER10, INTPTLAT10, INTPTLON10))

TXLA_svi <- svi %>%
  filter(GEOID %in% unlist(ZCTAs)) %>%
  select(-c(year, state)) %>%
  mutate(across(starts_with("E_"), as.integer))
```

```{r}
#| fig-height: 20

library(gridExtra)

output_maps <- colnames(TXLA_svi) %>% 
  `[`(str_detect(., "theme")) %>%
  map(~
    TXLA_svi %>% 
    left_join(ZCTAs_sf) %>%
    st_as_sf() %>%
    ggplot() +
      geom_sf(aes(fill = !!sym(.x)), linewidth = 0, color = NA) +
      coord_sf(datum = "ESRI:102003") +
      scale_fill_distiller(
        palette = "BuPu",
        guide = guide_colorbar(
          direction = "horizontal",
          title.position = "top")) +
      theme_void() +
      theme(
        plot.title = element_text(face = "bold", size = 16),
        plot.subtitle = element_text(face = "bold", size = 12),
        plot.caption = element_text(size = 10, hjust = 0),
        legend.title = element_text(face = "bold", size = 12),
        legend.text = element_text(face = "bold", size = 12),
        legend.title.align=0.5,
        legend.position = "bottom",
        legend.key.width = unit(dev.size()[1] / 20, "inches")) +
      labs(
        caption = "Author: Ryan Zomorrodi\nDate: 4/1/2024\nSource: FEMA National Flood Hazard Layer")
  )

do.call(grid.arrange, c(output_maps, ncol = 2))
```

Outputed SVI to parquet.
``` {r}
library(arrow)

TXLA_svi %>%
  write_parquet("output/TXLA_ZCTA_SVI1216.parquet")
```