Creating vectors and tables

Day 3

Felix May

Freie Universität Berlin @ Theoretical Ecology

Reproduce slides

library(dplyr)
library(tidyr)
library(readr)

Overview – Data management topics

What you learned so far

Reading data into R
Visualising data
Summarising data
Selecting rows or columns from tables

What you will learn today

Creating vectors and tables from scratch
Combining tables from different files
Adjusting the structure of tables ➔ tidy data

Why is this useful?

Preparing tables for data entry
- E.g., treatment values in experiments
Input data for model predictions (see days 5 - 9)
Adjust data for analysis or visualisation

Generating single vectors

Creating vectors in R

Any set of values

my_num <- c(0.05, 1, 2.5)
my_fac <- c("low", "high", "med")

Note

Vectors always include values from the same data type (i.e., numeric or categorical)

Generating ordered sequences

Ordered sequences with user-defined, start, end, and step size (argument by)

seq(from = 2, to = 4, by = 0.5)

[1] 2.0 2.5 3.0 3.5 4.0

Reminder: Function arguments can be specified by name and/or order

seq(2, 4, by = 0.5)

[1] 2.0 2.5 3.0 3.5 4.0

Choose length of sequence instead of step size
- by is calculated from length

seq(2, 4, length = 9)

[1] 2.00 2.25 2.50 2.75 3.00 3.25 3.50 3.75 4.00

Generating ordered sequences

Special short syntax with by = 1

seq(2, 6, by = 1)

[1] 2 3 4 5 6

2:6

[1] 2 3 4 5 6

Also step sizes of -1 works

5:-5

 [1]  5  4  3  2  1  0 -1 -2 -3 -4 -5

Vectors with repeated values

Helpful for observational or experimental data with replicates
Replicate a complete vector

rep(1:4, times = 3)

 [1] 1 2 3 4 1 2 3 4 1 2 3 4

Replicate each element a given number of time

rep(1:4, each = 3)

 [1] 1 1 1 2 2 2 3 3 3 4 4 4

Also works with categorical data

levels <- c("high", "medium", "low")
rep(levels, each = 4)

 [1] "high"   "high"   "high"   "high"   "medium" "medium" "medium" "medium"
 [9] "low"    "low"    "low"    "low"

Creating data tables

We use tibble() here
Similar to base::data.frame with some improvements

mydata <- tibble(x = 1:5, y = seq(100, 200, length = 5))
mydata

# A tibble: 5 × 2
      x     y
  <int> <dbl>
1     1   100
2     2   125
3     3   150
4     4   175
5     5   200

Creating data tables

Tables can include different data types

tibble(num = c(1,2,3),
       fac = c("a","b","c"),
       log = c(T,T,F))

# A tibble: 3 × 3
    num fac   log  
  <dbl> <chr> <lgl>
1     1 a     TRUE 
2     2 b     TRUE 
3     3 c     FALSE

Creating data tables

Single values are recycled to match longer vectors

tibble(x = 1:10, treatment = "control")

# A tibble: 10 × 2
       x treatment
   <int> <chr>    
 1     1 control  
 2     2 control  
 3     3 control  
 4     4 control  
 5     5 control  
 6     6 control  
 7     7 control  
 8     8 control  
 9     9 control  
10    10 control

Creating data tables

Functions for generating vectors can be called within tibble()

tibble(temp = rep(seq(2,4), times  = 2),
       treatment = rep(c("control","CO2"), each = 3))

# A tibble: 6 × 2
   temp treatment
  <int> <chr>    
1     2 control  
2     3 control  
3     4 control  
4     2 CO2      
5     3 CO2      
6     4 CO2

Reminder: Adding columns to tables

dat1 <- tibble(temp = rep(2:3, times  = 2),
               treatment = rep(c("control","CO2"), each = 2))

Using dplyr::mutate()

dat2 <- dat1 %>% mutate(tree_species = "pine")
dat2

# A tibble: 4 × 3
   temp treatment tree_species
  <int> <chr>     <chr>       
1     2 control   pine        
2     3 control   pine        
3     2 CO2       pine        
4     3 CO2       pine

Using base R

dat1$tree_species <- "pine"
dat1

# A tibble: 4 × 3
   temp treatment tree_species
  <int> <chr>     <chr>       
1     2 control   pine        
2     3 control   pine        
3     2 CO2       pine        
4     3 CO2       pine

Combining several tables

Preparation: Create several tables and files

Example: Data from different locations in different files

dat_pine <- dat1 %>% mutate(tree_species = "pine")
dat_beech <- dat1 %>% mutate(tree_species = "beech")
dat_oak <- dat1 %>% mutate(tree_species = "oak")

Write to csv

write_csv(dat_pine, file = "data/03_pine1.csv")
write_csv(dat_beech, file = "data/03_beech1.csv")
write_csv(dat_oak, file = "data/03_oak1.csv")

Combining several tables – by row

Reminder: reading tables from files
- Set your working directory correctly!

dat_pine <- read_csv("data/03_pine1.csv")
dat_beech <- read_csv("data/03_beech1.csv")
dat_oak <- read_csv("data/03_oak1.csv")

Put the tables together by row

bind_rows(dat_pine, dat_beech, dat_oak)

# A tibble: 12 × 3
    temp treatment tree_species
   <dbl> <chr>     <chr>       
 1     2 control   pine        
 2     3 control   pine        
 3     2 CO2       pine        
 4     3 CO2       pine        
 5     2 control   beech       
 6     3 control   beech       
 7     2 CO2       beech       
 8     3 CO2       beech       
 9     2 control   oak         
10     3 control   oak         
11     2 CO2       oak         
12     3 CO2       oak

Combining several tables – potential problems

Usually, the tables should have the same column names

names(dat_pine)

[1] "temp"         "treatment"    "tree_species"

dat_pine <- dat_pine %>% rename(Species = "tree_species")
names(dat_pine)

[1] "temp"      "treatment" "Species"

bind_rows(dat_pine, dat_beech)

# A tibble: 8 × 4
   temp treatment Species tree_species
  <dbl> <chr>     <chr>   <chr>       
1     2 control   pine    <NA>        
2     3 control   pine    <NA>        
3     2 CO2       pine    <NA>        
4     3 CO2       pine    <NA>        
5     2 control   <NA>    beech       
6     3 control   <NA>    beech       
7     2 CO2       <NA>    beech       
8     3 CO2       <NA>    beech

Combining several tables – solution

Mini-exercise

How would you fix the problem?

Solution code

dat_pine <- dat_pine %>% rename(tree_species = "Species")

bind_rows(dat_pine, dat_beech)

# A tibble: 8 × 3
   temp treatment tree_species
  <dbl> <chr>     <chr>       
1     2 control   pine        
2     3 control   pine        
3     2 CO2       pine        
4     3 CO2       pine        
5     2 control   beech       
6     3 control   beech       
7     2 CO2       beech       
8     3 CO2       beech

Combining tables – by column

dat2 <- tibble(year = 2023, 
               irrigation = rep(c("yes","no"), times = 2))
dat2

# A tibble: 4 × 2
   year irrigation
  <dbl> <chr>     
1  2023 yes       
2  2023 no        
3  2023 yes       
4  2023 no

Combining by columns

bind_cols(dat_pine, dat2)

# A tibble: 4 × 5
   temp treatment tree_species  year irrigation
  <dbl> <chr>     <chr>        <dbl> <chr>     
1     2 control   pine          2023 yes       
2     3 control   pine          2023 no        
3     2 CO2       pine          2023 yes       
4     3 CO2       pine          2023 no

Combining tables – potential problems

Tables must have equal row numbers (or just a single row)

dat3 <- tibble(year = 2023, 
               irrigation = rep(c("yes","no"), times = 3))
bind_cols(dat_pine, dat3)

Error in `bind_cols()`:
! Can't recycle `..1` (size 4) to match `..2` (size 6).

Exercise

Now it is your turn!

Exercise: Generating experimental treatments