R/repository.R

In antoinepay/foodx: Match your foods with Marmiton recipes

#' Fetch Food Data from OpenFoodFacts - French dataset
#' @import dplyr
#' @import readr
#' @import utils
#' @export
#'
fetch_data_from_openfoodfacts <- function() {
  
  food_csv_filepath <- 'data/food.csv'
  
  download.file('https://fr.openfoodfacts.org/data/fr.openfoodfacts.org.products.csv', food_csv_filepath)
  food <- read_tsv(food_csv_filepath)
  
  columns_to_remove <- c(
    'states',
    'states_tags',
    'states_fr',
    'url',
    'creator',
    'created_datetime', 
    'last_modified_datetime', 
    'image_small_url',
    'image_ingredients_url',
    'image_ingredients_small_url',
    'image_nutrition_small_url',
    'image_nutrition_url',"traces_tags",                                
    "traces_fr",                                  
    "serving_size",                               
    "traces",                                     
    "emb_codes",                                  
    "emb_codes_tags", 
    "cities_tags",
    "origins",
    "origins_tags",
    "created_t",
    "last_modified_t",
    "packaging",
    "packaging_tags",
    "manufacturing_places",
    "manufacturing_places_tags",
    "first_packaging_code_geo",
    "purchase_places","stores",
    "image_url",
    "image_ingredients_url",
    "image_ingredients_small_url", 
    "image_nutrition_url",
    "image_nutrition_small_url"
  )
  
  food <- food %>% 
    filter(nchar(code) == 13 & !is.na(product_name)) %>%
    select(-columns_to_remove) %>%
    filter(countries_tags=="en:france") %>% 
    select(which(colMeans(is.na(.)) < 0.8)) %>%
    filter_categories()
  
  write_csv(food, food_csv_filepath)
}

#' Get the best category of the product
#'
#' @param food_dataset data.frame from OpenFoodFacts 
#'
#' @return filtered categories dataset
#' @export
#'
filter_categories <- function(food_dataset) {
  food_dataset %>%  
    rowwise() %>% 
    mutate(
      categories = get_last_category(as.character(categories)),
      categories_tags = get_last_category(as.character(categories_tags)),
      categories_fr = get_last_category(as.character(categories_fr))
      )
}

#' Split categories and get the last item listed 
#'
#' @param categories character with multiple categories separated by a comma
#'
#' @return category with a single element
#'
get_last_category <- function(categories) {
  c <- strsplit(categories, ",")[[1]]
  c[[length(c)]]
}