Visualizing Data in the Tidyverse - Fast Food Restaurant Plots: Final Course Project
Giovanni Barrero Ortiz
4/3/2021
Introduction
we will create data visualizations using two data sets about fast food restaurants. the overall goal is to explore the nutrition of entree items and the sales of fast food restaurants in 2018. The task is to reconstruct the a set of plots, all of which were constructed using packages in the Tidyverse or packages that integrate with the Tidyverse. The source data is in : https://hubworks.com/blog/statistics-about-americas-biggest-fast-food-chains.html.
Question 1: Us Sales vs Total number of stores
Loading required package:
setwd("D:/coursera/TIDYVERSE/Visualizing Data in the Tidyverse/proyecto final")
library(tidyverse)## Warning: package 'tidyverse' was built under R version 4.0.4## -- Attaching packages --------------------------------------- tidyverse 1.3.0 --## v ggplot2 3.3.3 v purrr 0.3.4
## v tibble 3.0.5 v dplyr 1.0.3
## v tidyr 1.1.2 v stringr 1.4.0
## v readr 1.4.0 v forcats 0.5.0## -- Conflicts ------------------------------------------ tidyverse_conflicts() --
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()library(ggrepel)## Warning: package 'ggrepel' was built under R version 4.0.4Obtain the data:
sales<-read_csv("data_fastfood_sales.csv")##
## -- Column specification --------------------------------------------------------
## cols(
## restaurant = col_character(),
## average_sales = col_double(),
## us_sales = col_double(),
## num_company_stores = col_double(),
## num_franchised_stores = col_double(),
## unit_count = col_double()
## )cal <- read_csv("data_fastfood_calories.csv")##
## -- Column specification --------------------------------------------------------
## cols(
## restaurant = col_character(),
## item = col_character(),
## calories = col_double(),
## cal_fat = col_double(),
## total_fat = col_double(),
## sat_fat = col_double(),
## trans_fat = col_double(),
## cholesterol = col_double(),
## sodium = col_double(),
## total_carb = col_double(),
## fiber = col_double(),
## sugar = col_double(),
## protein = col_double(),
## vit_a = col_double(),
## vit_c = col_double(),
## calcium = col_double()
## )Summarizing the calories dataset
##install.packages("skimr")
library(skimr)## Warning: package 'skimr' was built under R version 4.0.4skim(cal)| Name | cal |
| Number of rows | 515 |
| Number of columns | 16 |
| _______________________ | |
| Column type frequency: | |
| character | 2 |
| numeric | 14 |
| ________________________ | |
| Group variables | None |
Variable type: character
| skim_variable | n_missing | complete_rate | min | max | empty | n_unique | whitespace |
|---|---|---|---|---|---|---|---|
| restaurant | 0 | 1 | 5 | 11 | 0 | 8 | 0 |
| item | 0 | 1 | 5 | 63 | 0 | 505 | 0 |
Variable type: numeric
| skim_variable | n_missing | complete_rate | mean | sd | p0 | p25 | p50 | p75 | p100 | hist |
|---|---|---|---|---|---|---|---|---|---|---|
| calories | 0 | 1.00 | 530.91 | 282.44 | 20 | 330.0 | 490.0 | 690 | 2430 | ▇▆▁▁▁ |
| cal_fat | 0 | 1.00 | 238.81 | 166.41 | 0 | 120.0 | 210.0 | 310 | 1270 | ▇▃▁▁▁ |
| total_fat | 0 | 1.00 | 26.59 | 18.41 | 0 | 14.0 | 23.0 | 35 | 141 | ▇▃▁▁▁ |
| sat_fat | 0 | 1.00 | 8.15 | 6.42 | 0 | 4.0 | 7.0 | 11 | 47 | ▇▃▁▁▁ |
| trans_fat | 0 | 1.00 | 0.47 | 0.84 | 0 | 0.0 | 0.0 | 1 | 8 | ▇▁▁▁▁ |
| cholesterol | 0 | 1.00 | 72.46 | 63.16 | 0 | 35.0 | 60.0 | 95 | 805 | ▇▁▁▁▁ |
| sodium | 0 | 1.00 | 1246.74 | 689.95 | 15 | 800.0 | 1110.0 | 1550 | 6080 | ▇▆▁▁▁ |
| total_carb | 0 | 1.00 | 45.66 | 24.88 | 0 | 28.5 | 44.0 | 57 | 156 | ▅▇▂▁▁ |
| fiber | 12 | 0.98 | 4.14 | 3.04 | 0 | 2.0 | 3.0 | 5 | 17 | ▇▅▂▁▁ |
| sugar | 0 | 1.00 | 7.26 | 6.76 | 0 | 3.0 | 6.0 | 9 | 87 | ▇▁▁▁▁ |
| protein | 1 | 1.00 | 27.89 | 17.68 | 1 | 16.0 | 24.5 | 36 | 186 | ▇▂▁▁▁ |
| vit_a | 214 | 0.58 | 18.86 | 31.38 | 0 | 4.0 | 10.0 | 20 | 180 | ▇▁▁▁▁ |
| vit_c | 210 | 0.59 | 20.17 | 30.59 | 0 | 4.0 | 10.0 | 30 | 400 | ▇▁▁▁▁ |
| calcium | 210 | 0.59 | 24.85 | 25.52 | 0 | 8.0 | 20.0 | 30 | 290 | ▇▁▁▁▁ |
Summarizing the sales dataset
skim(sales)| Name | sales |
| Number of rows | 19 |
| Number of columns | 6 |
| _______________________ | |
| Column type frequency: | |
| character | 1 |
| numeric | 5 |
| ________________________ | |
| Group variables | None |
Variable type: character
| skim_variable | n_missing | complete_rate | min | max | empty | n_unique | whitespace |
|---|---|---|---|---|---|---|---|
| restaurant | 0 | 1 | 3 | 15 | 0 | 19 | 0 |
Variable type: numeric
| skim_variable | n_missing | complete_rate | mean | sd | p0 | p25 | p50 | p75 | p100 | hist |
|---|---|---|---|---|---|---|---|---|---|---|
| average_sales | 0 | 1 | 1189.88 | 541.52 | 360.72 | 857.50 | 1130.00 | 1470.10 | 2670.32 | ▆▇▆▁▁ |
| us_sales | 0 | 1 | 7592.69 | 8007.32 | 606.00 | 3499.88 | 4476.41 | 9539.12 | 37480.67 | ▇▃▁▁▁ |
| num_company_stores | 0 | 1 | 839.00 | 1875.80 | 0.00 | 53.50 | 276.00 | 677.50 | 8222.00 | ▇▁▁▁▁ |
| num_franchised_stores | 0 | 1 | 5998.53 | 5894.51 | 0.00 | 2583.00 | 4055.00 | 6497.50 | 25908.00 | ▇▅▂▁▁ |
| unit_count | 0 | 1 | 6838.05 | 5997.13 | 2231.00 | 3034.50 | 4332.00 | 7394.00 | 25908.00 | ▇▁▂▁▁ |
Generating Plot
sales_w1<-sales %>%
mutate(prop=num_franchised_stores/unit_count)
p <- ggplot(sales_w1, aes(x=us_sales, y=unit_count)) +
geom_point(aes(color=prop)) +
scale_x_log10() +
scale_y_log10() +
theme_bw() +
geom_text_repel(aes(label=restaurant)) +
labs(x="Total number of stores (log10 scale)",
y="U.S. sales in millions (log10 scale)",
color="Proportion of stores franchised")
p
Question 2: Bar plot of average sales per restaurant
Using sales dataset above , create a bar plot with the average_sales on the x-axis and restaurant on the y-axis (Hint: consider using the coord_flip() function). The order of restaurants on the y-axis should be in decreasing order of average sales with the restaurant with the largest average sales at the top and the restaurant with the smallest average sales at the bottom. Add text to each bar on the plot with the average sales (in the thousands) for each restaurant. - Each axis should be appropriately labeled. Along the x-axis, transform the text labels to include a dollar sign in front of each number. Use the classic ggplot2 theme.
library(forcats)
library(scales)##
## Attaching package: 'scales'## The following object is masked from 'package:purrr':
##
## discard## The following object is masked from 'package:readr':
##
## col_factorq2 <- ggplot(sales, aes(x=average_sales, y=fct_reorder(restaurant, average_sales))) +
geom_bar(stat="identity") +
theme_classic() +
geom_text(aes(label=paste0("$", round(average_sales,0))), hjust=-0.1) +
scale_x_continuous(labels=dollar_format()) +
labs(x="Average sales per unit store (in thousands)",
y="Restaurant")
q2
Question 3: Calories vs Sodium
Relationship between calories and sodium. for different entrees and we will label entree items with more sodium than 2300 mg. Here create a scatter plot with the column calories along the xaxis and the column sodium along the y-axis. - Each restaurant should have its own scatter plot (Hint: consider the facet functions). - Add a horizontal line at y=2300 in each scatter plot. - Each axis of the scatter plot should have an appropriately labeled x-axis and yaxis. - For all food items with a sodium level of greater than 2300 (mg) (the maximum daily intake from the Centers for Disease Control), add a text label each point with the name of the entree food item using the ggrepel package. - Use the classic dark-on-light ggplot2 theme. - Rename the legend.
q3 <- ggplot(cal, aes(x=calories, y=sodium)) +
facet_wrap(~restaurant) +
geom_hline(yintercept=2300) +
theme_bw() +
labs(x="Calories", y="Sodium (mg)") +
geom_text_repel(data=cal %>% filter(sodium>2300),
aes(label=item),
direction="y",
nudge_y=750,
nudge_x=750,
size=3) +
geom_point(size=2)
q3## Warning: ggrepel: 2 unlabeled data points (too many overlaps). Consider
## increasing max.overlaps
Question 4: Calories vs Restaurant
Calories per entree at Fast Food Restaurants. Here use cal dataset above . Create a new column titled is_salad that contains a TRUE or FALSE value of whether or not the name of entree food item contains the character string “salad” in it. Create boxplots with calories on the x-axis and restaurant on the y-axis. - The order of restaurants on the y-axis should be in decreasing order of calories with the restaurant with the median calories at the top and the restaurant with the smallest median calories at the bottom. Hide any outliers in the boxplots. On top of the boxplots add a set of jittered points representing each food item. - Each point should be colored based on whether it is an item with the word “salad” in it or not. - Each axis should be appropriately labeled, the legend should be appropriately labeled, and the x-axis should be transformed to a log10 scale. Use the classic dark-on-light ggplot2 theme.
calories_q4<-cal %>%
mutate(is_salad=str_detect(item, "[Ss]alad"))
q4 <- ggplot(calories_q4, aes(x=fct_reorder(restaurant, calories), y=calories)) + geom_boxplot(outlier.shape = NA) +
geom_point(aes(color=is_salad),
position=position_jitterdodge()) +
scale_y_log10() +
coord_flip() +
theme_bw() +
labs(x="Restaurant", y="Calories (log10 scale)", color="Is the entree\n a salad?") +
scale_color_discrete(labels=c("Not a salad", "Salad"))
q4
Question 5: Restaurant vs Us Sales
Comparision of US Sales and Sugar Content at Fast Food Restaurants. Here use cal dataset above .Remove rows that contain the Taco Bell restaurant - For each restaurant calculate the median amount of sugar in each entree item. Using this summarized dataset, combine this summarized dataset with the data_fastfood_sales.csv dataset. The combined dataset should only include restaurants that are included in both datasets. Using this new dataset, create a bar plot with restaurant on the x-axis and on the us_sales y-axis. The order of restaurants on the x-axis should be in increasing order of US sales with the restaurant with the largest average sales on the right and the restaurant with the smallest US sales on the left. Color the bars by the median amount of sugar in the entree items from that restaurant. Each axis should be appropriately labeled. - Use the classic ggplot2 theme.
calories_q5<-cal %>%
filter(restaurant!="Taco Bell") %>%
group_by(restaurant) %>%
summarise(median_sugar=median(sugar)) %>%
filter(restaurant %in% sales$restaurant) %>%
inner_join(sales[,c(1,3)], by="restaurant")
ggplot(calories_q5, aes(x=fct_reorder(restaurant, us_sales), y=us_sales)) +
geom_bar(aes(fill=median_sugar), stat="identity") +
scale_fill_viridis_c() +
labs(x="Restaurant", y="U.S. sales (in millions)", fill="Median sugar (grams)\n in fast food
entrees") +
theme_classic()
Conclutions
In this project we have seen the power of graphics when it comes to displaying data analysis results. GGPLOT2 is a powerful and useful package to be used in data science.