Real World Example
Chicago has a population of about 2.7 million people, and
unfortunately, it has a higher than average crime rate, and all crime
data is contained in the dataset, found
here. In R, the chiOpenData package can be used to pull
this data directly.
By using the chi_pull_dataset() function, we can gather
the most recent crime cases in Chicago, and filter based upon any of the
columns inside the dataset.
Let’s take an example of 3 requests that occur on the street. The
chi_pull_dataset() function can filter based off any of the
columns in the dataset. To filter, we add filters = list()
and put whatever filters we would like inside. From our
colnames call before, we know that there is a column called
“location_description” which we can use to accomplish this.
chicago_crimes_street <- chi_pull_dataset(dataset = "ijzp-q8t2",limit = 3, timeout_sec = 90, filters = list(location_description = "STREET"))
chicago_crimes_street
#> # A tibble: 3 × 24
#> id case_number date block iucr primary_type description
#> <dbl> <chr> <dttm> <chr> <dbl> <chr> <chr>
#> 1 12131221 JD327000 2020-08-10 09:45:00 015XX… 326 ROBBERY AGGRAVATED…
#> 2 13061203 JG246126 2023-05-03 08:10:00 073XX… 486 BATTERY DOMESTIC B…
#> 3 13116982 JG312117 2023-06-23 04:44:00 040XX… 142 HOMICIDE RECKLESS H…
#> # ℹ 17 more variables: location_description <chr>, arrest <lgl>,
#> # domestic <lgl>, beat <dbl>, district <dbl>, ward <dbl>,
#> # community_area <dbl>, fbi_code <chr>, x_coordinate <dbl>,
#> # y_coordinate <dbl>, year <dbl>, updated_on <dttm>, latitude <dbl>,
#> # longitude <dbl>, location_latitude <dbl>, location_longitude <dbl>,
#> # location_human_address <chr>
# Checking to see the filtering worked
chicago_crimes_street |>
distinct(location_description)
#> # A tibble: 1 × 1
#> location_description
#> <chr>
#> 1 STREETSuccess! From calling the chicago_crimes_2026 dataset we
see there are only 3 rows of data, and from the distinct()
call we see the only location featured in our dataset is STREET.
One of the strongest qualities this function has is its ability to filter based off of multiple columns. Let’s put everything together and get a dataset of 50 crimes that occur on the STREET that are not domestic.
# Creating the dataset
chicago_crimes <- chi_pull_dataset(dataset = "ijzp-q8t2", limit = 50, timeout_sec = 90, filters = list(location_description = "STREET", domestic = FALSE))
# Calling head of our new dataset
chicago_crimes |>
slice_head(n = 6)
#> # A tibble: 6 × 24
#> id case_number date block iucr primary_type description
#> <dbl> <chr> <dttm> <chr> <chr> <chr> <chr>
#> 1 12131221 JD327000 2020-08-10 09:45:00 015XX… 0326 ROBBERY AGGRAVATED…
#> 2 13116982 JG312117 2023-06-23 04:44:00 040XX… 0142 HOMICIDE RECKLESS H…
#> 3 12888104 JF469015 2022-11-10 03:47:00 072XX… 1477 WEAPONS VIO… RECKLESS F…
#> 4 13129199 JG327622 2023-07-04 17:30:00 047XX… 0910 MOTOR VEHIC… AUTOMOBILE
#> 5 13179344 JG386917 2023-08-17 19:25:00 030XX… 0860 THEFT RETAIL THE…
#> 6 13183567 JG392073 2023-08-21 14:37:00 022XX… 1220 DECEPTIVE P… THEFT OF L…
#> # ℹ 17 more variables: location_description <chr>, arrest <lgl>,
#> # domestic <lgl>, beat <dbl>, district <dbl>, ward <dbl>,
#> # community_area <dbl>, fbi_code <chr>, x_coordinate <dbl>,
#> # y_coordinate <dbl>, year <dbl>, updated_on <dttm>, latitude <dbl>,
#> # longitude <dbl>, location_latitude <dbl>, location_longitude <dbl>,
#> # location_human_address <chr>
# Quick check to make sure our filtering worked
chicago_crimes |>
summarize(rows = n())
#> # A tibble: 1 × 1
#> rows
#> <int>
#> 1 50
chicago_crimes |>
distinct(location_description)
#> # A tibble: 1 × 1
#> location_description
#> <chr>
#> 1 STREET
chicago_crimes |>
distinct(domestic)
#> # A tibble: 1 × 1
#> domestic
#> <lgl>
#> 1 FALSEWe successfully created our dataset that contains 50 requests regarding that are not domestic that happen on the street.
Mini analysis
Now that we have successfully pulled the data and have it in R, let’s
do a mini analysis on using the primary_type column, to
figure out what are the main types of crimes.
To do this, we will create a bar graph of the crime types.
# Visualizing the distribution, ordered by frequency
chicago_crimes |>
count(primary_type) |>
ggplot(aes(
x = n,
y = reorder(primary_type, n)
)) +
geom_col(fill = "steelblue") +
theme_minimal() +
labs(
title = "Top 50 Crime Types on the Street That Are Not Domestic",
x = "Number of Crimes",
y = "Primary Crime Type"
)
Bar chart showing the frequency of crime types happening on the street that are not domestic.
This graph shows us not only which crimes were committed, but how many of each crime occurred. This suggests that theft is the most common crime type among recent non-domestic street incidents.