Module 2 Lab
McKellar
2026-01-25
Part 1
Introduction
Greetings - hope you’re doing well. I prefer to go by my last name, McKellar, instead of my first, although my friends shorten my name to Mick. I’m a nerd, a gamer, I work in IT, and in general like to play with puzzles and learn. I’m also a fan of stage musicals, which you can probably tell from the picture I’ve included here where I’m dressed as Patrick Star from the SpongeBob musical for my office Halloween costume contest.
Background
I got my bachelor’s in computer engineering from Rose-Hulman Institute of Technology, and now I’m ever so slowly getting an MBA from University of Cincinnati. In a perfect world I would have graduated already, but after enough curveballs from work and life that finish line has been pushed out a bit. Through my previous education and work I’ve worked with analytic software before, mainly Python Pandas and Power BI / Power Query.
Part 2
Part 2 Question 1
# import data and the message print out will also tell you
# what the data types are
df <- readr::read_csv("blood_transfusion.csv")## Rows: 748 Columns: 5
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (1): Class
## dbl (4): Recency, Frequency, Monetary, Time
##
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.There are the following number of missing values
# Are there any missing values?
sum(is.na(df))## [1] 0The data has the following dimensions
# What are the dimensions of this data
dim(df)## [1] 748 5With these values in the first and last 10 rows
# Check out the first 10 rows
head(df, 10)## # A tibble: 10 × 5
## Recency Frequency Monetary Time Class
## <dbl> <dbl> <dbl> <dbl> <chr>
## 1 2 50 12500 98 donated
## 2 0 13 3250 28 donated
## 3 1 16 4000 35 donated
## 4 2 20 5000 45 donated
## 5 1 24 6000 77 not donated
## 6 4 4 1000 4 not donated
## 7 2 7 1750 14 donated
## 8 1 12 3000 35 not donated
## 9 2 9 2250 22 donated
## 10 5 46 11500 98 donated# Check out the last 10 rows
tail(df, 10)## # A tibble: 10 × 5
## Recency Frequency Monetary Time Class
## <dbl> <dbl> <dbl> <dbl> <chr>
## 1 23 1 250 23 not donated
## 2 23 4 1000 52 not donated
## 3 23 1 250 23 not donated
## 4 23 7 1750 88 not donated
## 5 16 3 750 86 not donated
## 6 23 2 500 38 not donated
## 7 21 2 500 52 not donated
## 8 23 3 750 62 not donated
## 9 39 1 250 39 not donated
## 10 72 1 250 72 not donatedThe value in the ‘Monetary’ column at the 100th row is
# Index for the 100th row and just the `Monetary` column. What is the value?
df[100, 'Monetary']## # A tibble: 1 × 1
## Monetary
## <dbl>
## 1 1750The mean of the ‘Monetary’ column is
# Index for just the `Monetary` column. What is the mean of this vector?
mean(df[['Monetary']])## [1] 1378.676A subset of data where the value of the ‘Monetary’ column is less than the mean of that column
# Subset this data frame for all observations where `Monetary` is greater
# than the mean value. How many rows are in the resulting data frame?
above_avg <- df[['Monetary']] > mean(df[['Monetary']])
df[above_avg, 'Monetary']## # A tibble: 267 × 1
## Monetary
## <dbl>
## 1 12500
## 2 3250
## 3 4000
## 4 5000
## 5 6000
## 6 1750
## 7 3000
## 8 2250
## 9 11500
## 10 5750
## # ℹ 257 more rowsPart 2 Question 2
df <- readr::read_csv("PDI__Police_Data_Initiative__Crime_Incidents.csv")## Rows: 15155 Columns: 40
## ── Column specification ────────────────────────────────────────────────────────
## Delimiter: ","
## chr (34): INSTANCEID, INCIDENT_NO, DATE_REPORTED, DATE_FROM, DATE_TO, CLSD, ...
## dbl (6): UCR, LONGITUDE_X, LATITUDE_X, TOTALNUMBERVICTIMS, TOTALSUSPECTS, ZIP
##
## ℹ Use `spec()` to retrieve the full column specification for this data.
## ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.For the data provided in the dile ’PDI__Police_Data_Initiative__Crime_Incidents.csv’ it has the following dimensions
# dimensions of this data
dim(df)## [1] 15155 40This data has this amount of missing data and are in the following columns
# Are there any missing data
sum(is.na(df))## [1] 95592# If so, how many missing values are in each column?
colSums(is.na(df))## INSTANCEID INCIDENT_NO
## 0 0
## DATE_REPORTED DATE_FROM
## 0 2
## DATE_TO CLSD
## 9 545
## UCR DST
## 10 0
## BEAT OFFENSE
## 28 10
## LOCATION THEFT_CODE
## 2 10167
## FLOOR SIDE
## 14127 14120
## OPENING HATE_BIAS
## 14508 0
## DAYOFWEEK RPT_AREA
## 423 239
## CPD_NEIGHBORHOOD WEAPONS
## 249 5
## DATE_OF_CLEARANCE HOUR_FROM
## 2613 2
## HOUR_TO ADDRESS_X
## 9 148
## LONGITUDE_X LATITUDE_X
## 1714 1714
## VICTIM_AGE VICTIM_RACE
## 0 2192
## VICTIM_ETHNICITY VICTIM_GENDER
## 2192 2192
## SUSPECT_AGE SUSPECT_RACE
## 0 7082
## SUSPECT_ETHNICITY SUSPECT_GENDER
## 7082 7082
## TOTALNUMBERVICTIMS TOTALSUSPECTS
## 33 7082
## UCR_GROUP ZIP
## 10 1
## COMMUNITY_COUNCIL_NEIGHBORHOOD SNA_NEIGHBORHOOD
## 0 0The ‘DATE_REPORTED’ column has has the following range of dates
# Using the `DATE_REPORTED` column, what is the `range` of dates included in this data?
range(df[['DATE_REPORTED']])## [1] "01/01/2022 01:08:00 AM" "06/26/2022 12:50:00 AM"The column ‘SUSPECT_AGE’ has the following range
# Using `table()`, what is the most common age range for known `SUSPECT_AGE`s?
table(df[['SUSPECT_AGE']])##
## 18-25 26-30 31-40 41-50 51-60 61-70 OVER 70 UNDER 18
## 1778 1126 1525 659 298 121 16 629
## UNKNOWN
## 9003By using the function ‘table()’ we can see the number of incidents per zip code and them manipulate that data to present it from greatest number to fewest
# Use `table()` to get the number of incidents per zip code. Sort this table
# for those zip codes with the most activity to the least activity.
sort(table(df['ZIP']), decreasing = TRUE)## ZIP
## 45202 45205 45211 45238 45229 45219 45225 45214 45237 45223 45206 45220 45232
## 2049 1110 1094 956 913 863 811 774 699 653 616 477 477
## 45224 45209 45208 45204 45216 45227 45207 45203 45230 45213 45239 45226 45217
## 429 380 359 348 302 286 245 226 214 190 169 112 100
## 45221 45233 45212 45215 45231 45228 42502 45236 45244 45248 4523 5239
## 90 77 61 47 7 5 3 3 3 3 2 1Finally, using the same table method and some averaging we can see which day of the week the most incidents occur on
# Using the `DAYOFWEEK` column, which day do most incidents occur on? What
# is the proportion of incidents that fall on this day?
table(df[['DAYOFWEEK']]) / sum(table(df[['DAYOFWEEK']]))##
## FRIDAY MONDAY SATURDAY SUNDAY THURSDAY TUESDAY WEDNESDAY
## 0.1369807 0.1438365 0.1542221 0.1448547 0.1363019 0.1432935 0.1405105At a glance, three columns that jumped out to me are SUSPECT_AGE, WEAPONS, and HATE_BIAS, as it’d be interesting to see if there’s a trend or connection between these. Despite whatever interesting insights could be found from those values, there might not be enough SUSPECT_AGE data since of 15155 rows 9003 are marked as UNKNOWN. This can likely be exchanged for VICTIM_AGE, which while a different angle of analysis only has 2283 UNKNOWN values. The other columns are pretty solid without blanks, although even without blanks HATE_BIAS might not have enough data to yield interesting info since “N–NO BIAS/NOT APPLICABLE” makes up 13885 of its rows. Using these columns, I’d like to tabulate the average number of co-occurrences and see if anything jumps out.