Chiara Di Gravio
October 24, 2010
The data is taken from the Kaggle website.
The data contains nutrition facts from different food sources.
The 10 more represented countries in the data (Australia, Belgium, France, Germany, Italy, Portugal, Spain, Switzerland, United Kingdom and United States) are chosen and a detailed analysis is performed.
The code used to clean the data can be found here
dat <- read.csv("Intakes.csv", sep = "")
str(dat)
'data.frame': 891 obs. of 8 variables:
$ country : Factor w/ 10 levels "Australia","Belgium",..: 1 1 1 1 1 1 1 1 1 1 ...
$ main_category_en: Factor w/ 691 levels "Aliment-en-conserve",..: 3 5 6 7 8 14 17 18 22 23 ...
$ totEnergy : num 2400 239 301 1690 572 ...
$ totFat : num 37.3 3 4 0.5 1 2.5 0.1 0.5 17 15.1 ...
$ totSugar : num 47.1 3 3 90 24.3 0.5 94.9 4.3 1 0.6 ...
$ totSalt : num 0.1196 0.2921 0.3048 0.0229 9.4996 ...
$ meanScore : num 27 1 1 15 16 -6 22 -5 3 14 ...
$ healthy : Factor w/ 2 levels "healthy","unhealthy": 2 1 1 2 2 2 2 2 1 2 ...
The Shiny App allows the user to summarise intakes of the selected variable in the chosen country and to show whether the aforementioned variable is associated with the food being healthy or not.
The first panel (Summary) shows, the histogram, the summary statistics for the chosen variable as well as the dataset used for deriving both.
The second panel (Predict Healthy/Unhealthy Food) shows the results of a logistic regression where beign healthy/unhealthy is the outcome of interest and the selected variable is the only predictor. All the analyses are done by single country.