Count data divided by year and by region in R

Question

I have a very large (too big to open in Excel) biological dataset that looks something like this

    year <- c(1990, 1980, 1985, 1980, 1990, 1990, 1980, 1985, 1985,1990, 
              1980, 1985, 1980, 1990, 1990, 1980, 1985, 1985,
              1990, 1980, 1985, 1980, 1990, 1990, 1980, 1985, 1985)
    species <- c('A', 'A', 'B', 'B', 'B', 'C', 'C', 'C', 'A','A', 'A', 
                 'B', 'B', 'B', 'C', 'C', 'C', 'A', 'A', 'A', 'B', 'B', 'B', 
                 'C', 'C', 'C', 'A')
    region <- c(1, 1, 1, 3, 2, 3, 3, 2, 1, 1, 3, 3, 3, 2, 2, 1, 1, 1,1, 3, 3, 
                3, 2, 2, 1, 1, 1)
    df <- data.frame(year, species, region)

    df
    year species region
 1  1990       A      1
 2  1980       A      1
 3  1985       B      1
 4  1980       B      3
 5  1990       B      2
 6  1990       C      3
 7  1980       C      3
 8  1985       C      2
 9  1985       A      1
 10 1990       A      1
 11 1980       A      3
 12 1985       B      3
 13 1980       B      3
 14 1990       B      2
 15 1990       C      2
 16 1980       C      1
 17 1985       C      1
 18 1985       A      1
 19 1990       A      1
 20 1980       A      3
 21 1985       B      3
 22 1980       B      3
 23 1990       B      2
 24 1990       C      2
 25 1980       C      1
 26 1985       C      1
 27 1985       A      1

What I am looking to do is figure out how many of each species (A, B, or C) exist in each region (1, 2, or 3) in each of the three years I have (1980, 1985, or 1990).

I'm looking to end up with a dataset that looks something along the lines of this,

      region A_1980 B_1980 C_1980 A_1985 B_1985 C_1985 A_1990 B_1990 C_1990
 1      1      0      0      0      0      0      0      0      0      0
 2      2      1      1      1      1      1      1      1      1      1
 3      3      2      2      2      2      2      2      2      2      2

such that each row represents a region, and each column represents the count of each species, in a particular year. I've tried to do this using the spread function in conjunction with the group_by dplyr function, but I couldn't get it to do anything close to what I want.

Does anyone have any suggestions?

Tarsier

Answer 1

Something like this?

library(dplyr)

df2 <- df %>% 
  mutate(sp_year = paste(species, year, sep = "_")) %>%
  group_by(region) %>% 
  count(sp_year) %>% 
  spread(sp_year,n)

df2

Which gives this:

# A tibble: 3 x 10
# Groups:   region [3]
  region A_1980 A_1985 A_1990 B_1980 B_1985 B_1990 C_1980 C_1985 C_1990
   <dbl>  <int>  <int>  <int>  <int>  <int>  <int>  <int>  <int>  <int>
1      1      1      3      3     NA      1     NA      2      2     NA
2      2     NA     NA     NA     NA     NA      3     NA      1      2
3      3      2     NA     NA      3      2     NA      1     NA      1

Inebriate

also possible to use ?tidyr::unite instead of mutate(paste). Would be less verbose at the very least. – Internee

Answer 2

Similar to wl1234's answer but more concise. We can use unite to combine columns. We can also use count without group_by the variable. Finally, we can set fill = 0 in the spread function to replace NA with 0.

library(tidyverse)

df2 <- df %>%
  unite(sp_year, species, year, sep = "_") %>%
  count(sp_year, region) %>%
  spread(sp_year, n, fill = 0)
df2
# # A tibble: 3 x 10
#   region A_1980 A_1985 A_1990 B_1980 B_1985 B_1990 C_1980 C_1985 C_1990
#    <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>  <dbl>
# 1      1      1      3      3      0      1      0      2      2      0
# 2      2      0      0      0      0      0      3      0      1      2
# 3      3      2      0      0      3      2      0      1      0      1

Oram

This is awesome, and I love the NA => 0 addition as well! Thank you! – Tarsier

I didn't know about unite. I will use that instead of paste next time. – Inebriate