Challenge 3: Animal Weights

Tidy Data: Pivoting

Code

library(tidyverse)

knitr::opts_chunk$set(echo = TRUE, warning=FALSE, message=FALSE)

Read in data

For this challenge, I’ll read in the csv file on animal weights.

Code

library(readr)
animal_weight <- read_csv("_data/animal_weight.csv", show_col_types =  FALSE)
head(animal_weight)

# A tibble: 6 × 17
  IPCC A…¹ Cattl…² Cattl…³ Buffa…⁴ Swine…⁵ Swine…⁶ Chick…⁷ Chick…⁸ Ducks Turkeys
  <chr>      <dbl>   <dbl>   <dbl>   <dbl>   <dbl>   <dbl>   <dbl> <dbl>   <dbl>
1 Indian …     275     110     295      28      28     0.9     1.8   2.7     6.8
2 Eastern…     550     391     380      50     180     0.9     1.8   2.7     6.8
3 Africa       275     173     380      28      28     0.9     1.8   2.7     6.8
4 Oceania      500     330     380      45     180     0.9     1.8   2.7     6.8
5 Western…     600     420     380      50     198     0.9     1.8   2.7     6.8
6 Latin A…     400     305     380      28      28     0.9     1.8   2.7     6.8
# … with 7 more variables: Sheep <dbl>, Goats <dbl>, Horses <dbl>, Asses <dbl>,
#   Mules <dbl>, Camels <dbl>, Llamas <dbl>, and abbreviated variable names
#   ¹​`IPCC Area`, ²​`Cattle - dairy`, ³​`Cattle - non-dairy`, ⁴​Buffaloes,
#   ⁵​`Swine - market`, ⁶​`Swine - breeding`, ⁷​`Chicken - Broilers`,
#   ⁸​`Chicken - Layers`

Briefly describe the data

Code

dim(animal_weight)

[1]  9 17

Code

colnames(animal_weight)

 [1] "IPCC Area"          "Cattle - dairy"     "Cattle - non-dairy"
 [4] "Buffaloes"          "Swine - market"     "Swine - breeding"  
 [7] "Chicken - Broilers" "Chicken - Layers"   "Ducks"             
[10] "Turkeys"            "Sheep"              "Goats"             
[13] "Horses"             "Asses"              "Mules"             
[16] "Camels"             "Llamas"            

Code

unique(animal_weight$`IPCC Area`)

[1] "Indian Subcontinent" "Eastern Europe"      "Africa"             
[4] "Oceania"             "Western Europe"      "Latin America"      
[7] "Asia"                "Middle east"         "Northern America"   

The data set shows the weights of 16 different animal types (for agricultural use) for 9 different geographic regions. There are 16 different types of animal (including different uses for the same animal - such as chickens meant for broiling and chickens meant for laying eggs or dairy/nondairy cattle). The data set should be flipped so that instead of one long row giving 16 different weights for one geographic region, we only see one observation (animal weight) per row.

Anticipate the End Result

We’ll use pivot_longer() to list each of the 16 animal types under a column called Animals and their weights under a column called Weight instead. This should result in a much longer version of our current tibble, where each IPCC area (geographic region) will appear 16 times - once per different type of animal - instead of once. As there are 9 IPCC areas listed now, this should end up being a tibble of 144 rows (observations) and 3 columns (IPCC area, Animals, and Weight).

Pivot the Data

Now we will pivot all of the columns describing animal weights (columns 2 through 17) so that the name of each of these columns (Cattle - dairy through Llamas) is contained under a new column titled Animals. The values listed under the original columns will move to another new column, titled Weight.

Code

animal_weight<-pivot_longer(animal_weight, col = c(2:17),
                 names_to="Animals",
                 values_to = "Weight")
animal_weight

# A tibble: 144 × 3
   `IPCC Area`         Animals            Weight
   <chr>               <chr>               <dbl>
 1 Indian Subcontinent Cattle - dairy      275  
 2 Indian Subcontinent Cattle - non-dairy  110  
 3 Indian Subcontinent Buffaloes           295  
 4 Indian Subcontinent Swine - market       28  
 5 Indian Subcontinent Swine - breeding     28  
 6 Indian Subcontinent Chicken - Broilers    0.9
 7 Indian Subcontinent Chicken - Layers      1.8
 8 Indian Subcontinent Ducks                 2.7
 9 Indian Subcontinent Turkeys               6.8
10 Indian Subcontinent Sheep                28  
# … with 134 more rows

Check Dimensions of Pivoted Data

Before pivoting, I calculated that the new dimensions would be 144 rows and 3 columns, where each row describes the weight of a type of agricultural animal in a geographic region.

Code

dim(animal_weight)

[1] 144   3

Now that the data has been pivoted long, our resulting tibble dimensions are \(144x3\) as expected.