02-gis.Rmd

---
title: "Working with Maps and GIS data"
#author: "Liming Wang"
#date: "8/21/2017"
output: 
  html_document:
    toc: false
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## Notable packages

### The tidycensus packages
- Website: https://walkerke.github.io/tidycensus/
- Install with `install.packages("tidycensus")`
- Vignettes: https://walkerke.github.io/tidycensus/articles/basic-usage.html
- Reference: https://walkerke.github.io/tidycensus/reference/index.html

### The simple feature packages
- Website: https://github.com/r-spatial/sf
- Install with `install.packages("sf")`
    - On Mac, you may need to install `udunits` first, for example, with [homebrew](https://stackoverflow.com/a/42387825/688693)
    - On Linux, install `libudunits2-dev` with your software management system first
- Vignettes: https://r-spatial.github.io/sf/articles/sf1.html
- Reference: https://r-spatial.github.io/sf/reference/index.html


## Code examples

### Data files used in the examples 

1. [Portland Metro 1994 TAZ shape file](data/taz1260.shp.zip): Download & unzip to `data` subdirectory of your RStudio project folder
2. [Portland Metro 1994 Survey geocode.raw.zip](data/portland94_geocode.raw.zip)

### Map ACS 2012-16 Median Household Income by Census Tract

```{r}
## Install the tidycensus package if you haven't yet
#install.packages("tidycensus")

library(tidycensus)
library(ggplot2)
library(dplyr)

## setup cenus api key
## signup your census api key at http://api.census.gov/data/key_signup.html
#census_api_key("YOUR_CENSUS_API_KEY") # 
portland_tract_medhhinc <- get_acs(geography = "tract", 
                        year = 2016, # 2012-2016
                        variables = "B19013_001",  # Median Household Income in the Past 12 Months
                        state = "OR", 
                        county = c("Multnomah County", "Washington County", "Clackamas County"),
                        geometry = TRUE) # load geometry/gis info

(myplot <- ggplot(portland_tract_medhhinc) + 
  geom_sf(aes(fill = estimate)) +
  coord_sf(datum = NA) + theme_minimal())

ggsave("output/mymap.pdf", myplot)

```

### Interactive Maps of ACS 2012-16 Median Household Income by Census Tract

```{r, results="asis"}
## Install the mapview package if you haven't yet
#install.packages("mapview")
library(sf)
library(mapview)
library(dplyr)

mapview(portland_tract_medhhinc %>% select(estimate), 
        col.regions = sf.colors(10), alpha = 0.1)
```

### Example of spatial analysis: spatial join

```{r}
library(sf)
library(readr)
# read 1994 Metro TAZ shape file
taz_sf <- st_read("data/taz1260.shp", crs=2913)

# read geocode.raw file that contains X and Y coordinates
portland94_df <- read_csv("data/portland94_geocode.raw.zip", col_names=c("uid", "X", "Y", "case_id", 
                                                                       "freq", "rtz", "sid", 
                                                                       "totemp94", "retemp94"))
portland94_df <- portland94_df %>% 
  filter(X!=0, Y!=0) %>% 
  sample_n(500)

# create a point geometry with x and y coordinates in the data frame
portland94_sf <- st_as_sf(portland94_df, coords = c("X", "Y"), crs = 2913)

# spatial join to get TAZ for observations in portland94_sf
portland94_sf <- st_join(portland94_sf, taz_sf)
head(portland94_sf)

ggplot() +
  geom_sf(data = taz_sf, aes(alpha=0.9)) +
  geom_sf(data = portland94_sf)
```

# Resources:
- [R spatial CRAN Task View](https://cran.r-project.org/web/views/Spatial.html)