Classroom Instructions for 3/27/2017 - Correlations and Scatterplots

03272017.txt

## RUN ALL THIS SYNTAX BEFORE WE START
## This will install and load all the packages you need for class today. 

ipak <- function(pkg){
    new.pkg <- pkg[!(pkg %in% installed.packages()[, "Package"])]
    if (length(new.pkg)) 
        install.packages(new.pkg, dependencies = TRUE)
    sapply(pkg, require, character.only = TRUE)
}

packages <- c("ggplot2", "dplyr", "car", "highcharter", "ggcorrplot")
ipak(packages)

## STOP AND WAIT
## Take a look at this dataset

head(mtcars)

### What is the relationship between mpg and wt? 
### Here's link for how to interpret correlation coefficients: http://i.imgur.com/OtwLQH6.png
### Now visualize that

+ geom_point() Will create a scatterplot of your values
+ geom_smooth() Will add a trend line to your values
+ geom_smooth(method = lm) Will make that line straight

### Load our height and weight data. 
pop <- read.csv("https://raw.githubusercontent.com/ryanburge/pls2003_sp17/master/population.csv")

What is the relationship between height and weight in the general population? 

### Read is some syntax, tell me what you think it does. Run each line and see what you get. 

team <- read.csv("https://raw.githubusercontent.com/ryanburge/pls2003_sp17/master/teamstats.csv")
bball <- read.csv(url("https://raw.githubusercontent.com/ryanburge/pls2003_sp17/master/dplyr.csv"))

sal <- bball %>% group_by(team_id, year) %>% summarise(salary = mean(salary))

### Now, run this line: 

merge <- merge(team, sal, by=c("team_id", "year"))

## Let's answer some questions. 
## What is the relationship between how much you pay your players and how much your team wins? 
## Can you tell me the difference between AL and NL teams? 
## Is there a relationship between salaries paid and total attendance?

## Let's say we wanted to look for a whole bunch of correlations at one time. 
## We use the select command to create pick out some columns that could be related.

bbcor <- select(merge, w, l, r, ab, h, double, triple, hr, bb, so, sb, cs, attendance, salary)

## Then we create a new dataset that just contains correlation coefficients, rounded to one decimal place

cor <- round(cor(bbcor), 1)

## Then we plot!

ggcorrplot(cor, hc.order = TRUE, 
           type = "lower", 
           lab = TRUE, 
           lab_size = 3, 
           method="circle", 
           colors = c("tomato2", "white", "springgreen3"), 
           title="Correlogram of Baseball Stats", 
           ggtheme=theme_bw)
           
## What do you see? What is related?