R Markdown

I recommend using R Markdown when completing assignments. Information is available online at http://rmarkdown.rstudio.com.

In R Markdown you can embed R code. Here is a version of our Week 1 data exploration, but formatted nicely in Markdown.

Loading and Verifying data

Load data and packages. Notice that in Markdown I have to use two dots, not one.

library(tidyverse)
## Loading tidyverse: ggplot2
## Loading tidyverse: tibble
## Loading tidyverse: tidyr
## Loading tidyverse: readr
## Loading tidyverse: purrr
## Loading tidyverse: dplyr
## Conflicts with tidy packages ----------------------------------------------
## filter(): dplyr, stats
## lag():    dplyr, stats
IntroData <- read.csv("../data/6003-1-introdata.csv")

Check whether it loaded correctly.

head(IntroData)
##   fishid         species length_cm
## 1      1 atlantic_salmon        41
## 2      2 atlantic_salmon        38
## 3      3 atlantic_salmon        43
## 4      4 atlantic_salmon        46
## 5      5 atlantic_salmon        31
## 6      6 atlantic_salmon        50
str(IntroData)
## 'data.frame':    20 obs. of  3 variables:
##  $ fishid   : int  1 2 3 4 5 6 7 8 9 10 ...
##  $ species  : Factor w/ 2 levels "atlantic_salmon",..: 1 1 1 1 1 1 1 1 1 1 ...
##  $ length_cm: int  41 38 43 46 31 50 49 47 45 46 ...

The variables are:

Let’s verify that each one looks like it’s error-free.

fishid

This should be a number with one unique value per fish. str() above told us it’s an integer.

Let’s look at the range of values, and check for duplicates:

range(IntroData$fishid) 
## [1]  1 20
IntroData %>% # Do stuff to IntroData
  group_by(fishid) %>% # 1. Group by the fish id
  filter(n() > 1) # 2. Return anything that occurs more than once
## # A tibble: 0 x 3
## # Groups:   fishid [0]
## # ... with 3 variables: fishid <int>, species <fctr>, length_cm <int>

No duplicates. All good.

species

There should just be two levels of species. Is that true?

levels(IntroData$species) 
## [1] "atlantic_salmon" "porbeagle_shark"

Yep, all good. Moving on.

length_cm

Let’s make sure the total lengths are okay. Plot:

p <- ggplot(IntroData, aes(x = species, y = length_cm)) 

p + geom_jitter()

Are these values reasonable? If you collected the data yourself you’d know - but you didn’t do that here (I gave you the data myself).

One option is to trust me. But if you’re suspicious, go to http://fishbase.org/ and look up these fish species. Never hurts to check!

Analysis

Not much we can do from these data, aside from some very basic visualizations.

Let’s make a boxplot

Nice. Now, a histogram:

Here, I’ve used echo=FALSE so you don’t see my R code. You can check the Rmd file for it if you want.