Setup

• Checkout the etherpad for today https://etherpad.wikimedia.org/p/cIVbE45bGm

• Install/load the following packages:

library(tidyr)
library(dplyr)

Slides from today

• data_mashups.pptx

Download the following datasets:

1. mammal_count_data.csv

2. mammal_site_data.csv

3. mammal_physiology_data.csv

Time to read these data files into R.

# We'll have to make several tweaks here. 
counts <- read.csv('datasets/mammal_count_data.csv')
sites  <- read.csv('datasets/mammal_site_data.csv')
phys   <- read.csv('datasets/mammal_physiology_data.csv')

Post reading data warm-up option 1:

• Let’s pretend that small mammals are counted in quadrats like plants. Use dplyr to calculate a new column that contains the density of small mammals given that our quadrat size was 2 x 2 m.

Post reading data warm-up option 2:

• Using tidyr, expand our dataframe to include all species counts (i.e., zero if a species wasn’t observed) at a site.

---

Data mashups!

You are often going to collect some data that you then want to join up data from multiple tables/data.frames.

• Talk about unique keys.

• Make unique key for lookup!

• Joining A with B

• left_join (matches B onto A)

• right_join (matches A onto B)

• inner_join (retain only matching values in A and B)

• full_join (retain all values in all rows)

• anti_join (filtering join, no actual joining - get all rows in A, not found in B)

• Concept of one to many