1.1 Loop through State Folders

First, we start with the outer loop, which goes through all of the state folders, as defined above. Once in each state folder, we’ll identify all of the files that exist within it (usually a few .csv files and one .md file).

for (state_ in states) {
  # Within each data folder, there are multiple .csv files, one for each election. Let's pull out each of these files
  state_contents <- list.files(paste(here("Data/Polling-Places-Data/"), state_, sep = "/"))

1.2 Loop through CSV Files

Next, let’s loop through each of the states to read our .csv files in. Remember that in this step we will:
- Determine if the file is .csv (otherwise, it’s .md)
- Determine the name of the file (a combination of our file path, the state name from the outer loop, and the value from the inner loop)
- Read the file in as a temporary dataframe

  for (i in state_contents) {
    # If the file is a csv file, then let's read it in
    if (str_detect(i, ".csv") ) {
      # Find the file path name for the csv
      csv_name <- paste("Data/Polling-Places-Data", state_, i, sep = "/")

      # Let's read in the file as a temporary dataframe
      temp_df <- read_csv(csv_name, col_types = cols(
        .default = col_character(),
        election_date = col_date(format = ""),
        state = col_character(),
        jurisdiction = col_character(),
        county_name = col_character(),
        municipality = col_character(),
        jurisdiction_type = col_character(),
        polling_place_type = col_character(),
        precinct_id = col_character(),
        precinct_name = col_character(),
        name = col_character(),
        address = col_character()
        )
      )

1.3 Create New Features

Now that we’ve read in our dataset, we’ll create two new columns in the temporary dataframe:
1. state_name | From the outer loop
2. year | From the name of the csv (also available as the iterator, i)

  # We want the state name from earlier to be brought in as the column
  temp_df <- temp_df %>%
    mutate(state_name = state_,
           year = as.integer(str_extract(i, "\\d+"))
    )

1.4 Union to Master Dataframe

Lastly, we’ll append the temporary dataframe to our master dataframe using a smart union (bind_rows), which doesn’t need all column headers to match.

  # Lastly, let's append this to the bottom of our master dataset
  polling_places <- bind_rows(polling_places, temp_df)

Altogether Now!

What does all of this look like together? Well, here it is:

for (state_ in states) {
  # Within each data folder, there are multiple .csv files, one for each election. Let's pull out each of these files
  state_contents <- list.files(paste(here("Data/Polling-Places-Data/"), state_, sep = "/"))

  # Let's loop through each of the states to read and add our .csv files in
  for (i in state_contents) {
    # If the file is a csv file, then let's read it in
    if (str_detect(i, ".csv") ) {
      # print("Parsing...")

      # Find the file path name for the csv
      csv_name <- paste("Data", "Polling-Places-Data", state_, i, sep = "/")

      # Let's read in the file as a temporary dataframe
      temp_df <- read_csv(csv_name, col_types = cols(
        .default = col_character(),
        election_date = col_date(format = ""),
        state = col_character(),
        jurisdiction = col_character(),
        county_name = col_character(),
        municipality = col_character(),
        jurisdiction_type = col_character(),
        polling_place_type = col_character(),
        precinct_id = col_character(),
        precinct_name = col_character(),
        name = col_character(),
        address = col_character()
        )
      )

      # We want the state name from earlier to be brought in as the column
      temp_df <- temp_df %>%
        mutate(state_name = state_,
               year = as.integer(str_extract(i, "\\d+"))
        )

      # Lastly, let's append this to the bottom of our master dataset
      polling_places <- bind_rows(polling_places, temp_df)
    }
  }
  # cat("===========================\n===========================\n")
}

# WE'RE DONE.
cat(praise(), "WE'RE DONE!!!")

## You are slick! WE'RE DONE!!!

And that’s it! If you run the code above, make sure to uncomment the print/cat functions so you can get updates as the code runs. Let’s take a look at the final dataset:

Looks great!