Accelerated development

Accelerated development

Imperative

library(lme4)

d <- read.csv("my.file.csv")
#...
#...
o1 <- lmer(d$d1 ~ d$d2 + (1|d$d3))
#...
#...
plot(d$d2, d$d1)

Named

library(lme4)

flu.data <- read.csv("my.file.csv")
#...
#...
res <- lmer(antigenic.distance ~ num.mutation + (1 | date), data = flu.data)
#...
#...
plot(antigenic.distance ~ num.mutation, data = flu.data)

Commented

library(lme4)

flu.data <- read.csv("my.file.csv")
# Some light processing to get data into correct format
#...
#...
# Now the analyses – a linear mixed effect model
res <- lmer(antigenic.distance ~ num.mutation + (1 | date), data = flu.data)
# Checking the model is a good one
#...
#...
# Phew, it is, so plot the best explanatory and response variables!
plot(antigenic.distance ~ num.mutation, data = flu.data)

Structured

library(lme4)

### <b>
# Function to load in data from file and process
read_flu_data <- function(filename) {
  #' Load in data from file and process
  
  data <- read.csv(filename)
  # Some light processing to get into correct format
  #...
  data
}
### </b>

flu.data <- read_flu_data("my.file.csv")
# Now the analyses – a linear mixed effect model
res <- lmer(antigenic.distance ~ num.mutation + (1 | date), data = flu.data)
# Checking the model is a good one
#...
#...
# Phew, it is, so plot the best explanatory and response variables!
plot(antigenic.distance ~ num.mutation, data = flu.data)

More structured

helper.R

# All of the helper functions for our 
# experiments
library(lme4)

### <b>
read_flu_data <- function(filename) {
  # Load data and process it 
  data <- read.csv(filename)
  
  # Wrangle data
  #...

  data
}
### </b>

### <b>
check_flu_model <- function(model.out) {
  # Check model
  is.good.model <- #...
    
  #...
  is.good.model
}
### </b>

script.R

# Load in our generic helper functions
source("helper.R")

# Load data and process it 
my.data <- read_flu_data("my.file.csv")

# Analyses - linear mixed effect model
res <- lmer(antigenic.distance ~ 
              num.mutation +
              (1 | date), 
            data = flu.data)

# Check model
if (!check_flu_model(res))
  stop("Model is rubbish, give up now!")

# Plot the best explanatory and
# response variables
plot(antigenic.distance ~ num.mutation,
     data = flu.data)