Week 8

Quick hands-on

1. Bring in census data, tidily and easily

There’s a great R package for importing census data. Let’s install and load it.

library(leaflet)
library(tidyverse)

install.packages("tidycensus")
library(tidycensus)

To get the data from the Census’ API, you need to provide an API Key. Find yours and use it.

census_api_key("TK", overwrite = FALSE, install = FALSE)

Once you’ve done that, you can quickly grab data. If you want to know the median rent by state in 1990 … now you can.

m90 <- get_decennial(geography = "state", variables = "H043A001", year = 1990)

It can also easily be plotted.

m90 %>%
  ggplot(aes(x = value, y = reorder(NAME, value))) +
  geom_point()

We can also grab data from the more-frequently-updated American Community Survey. It’s like a rolling Census that’s being conducted all the time.

Let’s look at public transit.

transpo <- get_acs(geography = "state", variables = "B08006_008", geometry = FALSE, survey = "acs5", year = 2017)

head(transpo)

Interesting. But what we really want is the rate of transit riders. Not the raw number. So let’s get that, starting with the total number of folks who commute to work.

transpo_total <- get_acs(geography = "state", variables = "B08006_001", geometry = FALSE, survey = "acs5", year = 2017)
head(transpo_total)

Alright, now we need to get these … together in the same data frame! That’s where the join comes in. We did some of these in QGIS. What could we use in these two datasets to link them up?

2. A detour into joins

All you need to join is a common column. Here’s how it works in tidyverse.

transpo <- transpo %>% left_join(transpo_total, by = "NAME")

Now we can get the rate.

transpo$rate <- transpo$estimate.x / transpo$estimate.y * 100
head(transpo)

3. Map out (after another join)

We need to join that transportation data to our shapefile. Unfortunately, the syntax there is a little different. Fortunately, it does work.

For this, we’ll need to bring back the states shapefile we used last week,

library(rgdal)
states <- readOGR("path/to/yourfile/",
  layer = "tl_2019_us_state", GDAL1_integer64_policy = TRUE)

Then we can do a join.

states_with_rate <- sp::merge(states, transpo, by = "NAME")

Let’s try this out.

qpal <- colorQuantile("PiYG", states_with_rate$rate, 9)


states_with_rate %>% leaflet() %>% addTiles() %>%
  addPolygons(weight = 1, smoothFactor = 0.5, opacity = 1.0, fillOpacity = 0.5,
    color = ~qpal(rate),
    highlightOptions = highlightOptions(color = "white", weight = 2,
      bringToFront = TRUE))

Alright. What does each line do? Let’s play around with it and see what changes?

If we have extra time, we’ll work on adding popup text and a legend.

Homework

Github: Create an account if you don’t already have one. If you do, make sure you know the password.
HW: Finish coding if we didn’t in class.
Final Project: Your data analysis should almost done. You should be reporting and writing as you go.
Story memo: 50-100 words about Final Project progress over last week

Data Journalism USC 2020

Week 8

Quick hands-on

Homework