Download and open the assign04.qmd file and complete the exercises.
We will be using pay-per-click (PPC) data from a 31 day campaign from a company that sells USB keys and USB hubs. Each row of the 555 observations represents a click on an internet ad based on a keyword search and there are 3 columns.
day - represents the day of the campaign. Valid days are 1-31.
price - represents the price of the campaign. Price can’t be a number below 0.10
keyword - represents the keyword purchased. Everything must be spelled correctly, there aren’t many keywords but they are some combination of “usb” and/or “key” or “hub”
In this assignment you will be examining each column for data validity. Each exercise presents one or more questions for you to answer.
We’ll start by loading the tidyverse family of packages along with the janitor and skimr packages, and our data. Make sure you install these two packages in your RStudio prior to calling the library() functions below.
library(tidyverse)
library(janitor)Warning: package 'janitor' was built under R version 4.4.3library(skimr)Warning: package 'skimr' was built under R version 4.4.3ppc_data <- read_csv("https://jsuleiman.com/datasets/ppc_data.csv")
view(ppc_data)There are six exercises in this assignment. The Grading Rubric is available at the end of this document.
Create a graph of number of clicks (i.e., observations) for each day (1-31). Use geom_bar() for your geometry. In the narrative below your code note which days had zero clicks.
library(ggplot2)
ppc_data |>
ggplot(aes(x = day)) +
geom_bar(stat = "count") +
labs(title = "Number of Clicks per Day", x = "Day of the Month", y = "Number of Clicks") +
theme_minimal()
Day 27 Had 0 clicks
Insert a code cell to show how many NA (i.e., missing) values there are in price. In the narrative below that code cell write out how many NA values there are for price and what percent of the observations that represents.
ppc_data |>
select(price) |>
filter(is.na(price)) # A tibble: 6 × 1
price
<dbl>
1 NA
2 NA
3 NA
4 NA
5 NA
6 NAThere are 6 NA values under price, and with there being 555 observations that would mean that it comes out to about 1.08% 6/555 = 0.01081 and then multiply by 100 to get 1.08%
Valid values for price are 0.1 or greater. Insert a code cell that displays the number of values of price that are less than 0.1. In the narrative below that code cell write how many values are below 0.1.
ppc_data |>
select(price) |>
filter(price < 0.1) # A tibble: 10 × 1
price
<dbl>
1 0
2 0
3 0
4 0
5 0
6 0
7 0
8 0
9 0
10 0Insert a code cell that drops all of the rows that contain invalid or NA values for price.
ppc_data |>
select(price) |>
filter(is.na(price)) # A tibble: 6 × 1
price
<dbl>
1 NA
2 NA
3 NA
4 NA
5 NA
6 NAInsert a code cell that shows a tabyl of the counts of each keyword. In the narrative below the code cell, list the misspellings and counts if there are any.
keyword_counts <- ppc_data |>
tabyl(keyword)
glimpse(keyword_counts)Rows: 7
Columns: 3
$ keyword <chr> "hub", "key", "ubs", "ubs key", "usb", "usb hub", "usb key"
$ n <int> 90, 85, 11, 10, 75, 121, 163
$ percent <dbl> 0.16216216, 0.15315315, 0.01981982, 0.01801802, 0.13513514, 0.…ubs instead of usb 11 times
ubs key instead of usb key 10 times
Insert a code cell that corrects all the misspellings for keyword, then rerun tabyl to verify.
ppc_data |>
mutate(keyword = recode(keyword, "ubs" = "usb", "ubs key" = "usb key")) |>
tabyl(keyword) keyword n percent
hub 90 0.1621622
key 85 0.1531532
usb 86 0.1549550
usb hub 121 0.2180180
usb key 173 0.3117117To submit your assignment:
| Item (percent overall) |
100% - flawless | 67% - minor issues | 33% - moderate issues | 0% - major issues or not attempted |
|---|---|---|---|---|
| Narrative: typos and grammatical errors (7%) |
||||
| Document formatting: correctly implemented instructions (7%) |
||||
Exercises (13% each) |
||||
Submitted properly to Brightspace (8%) |
NA | NA | You must submit according to instructions to receive any credit for this portion. |