install.packages("ggplot2", repos = 'https://cran.rstudio.com/bin/macosx/mavericks/contrib/3.3/ggplot2_2.1.0.tgz')
## Warning: unable to access index for repository https://cran.rstudio.com/bin/macosx/mavericks/contrib/3.3/ggplot2_2.1.0.tgz/src/contrib:
##   cannot download all files
## Warning: package 'ggplot2' is not available (for R version 3.3.1)
## Warning: unable to access index for repository https://cran.rstudio.com/bin/macosx/mavericks/contrib/3.3/ggplot2_2.1.0.tgz/bin/macosx/mavericks/contrib/3.3:
##   cannot download all files
library(ggplot2)

data(diamonds)
pf <- read.csv('pseudo_facebook.tsv', sep = '\t')

Price Histograms with Facet and Color

ggplot(aes(x = price, color = cut), data = diamonds) +
  facet_wrap(~color, ncol = 3) + 
  geom_histogram() + 
  scale_x_log10() + 
  scale_fill_brewer(type = 'qual')
## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

Price vs. Table Colored by Cut

ggplot(data = diamonds, aes(x = table, y = price, color = cut)) + 
  geom_point(alpha = 1/5) + 
  scale_x_continuous(limits = c(50, 80), breaks = seq(50, 80, 2))
## Warning: Removed 5 rows containing missing values (geom_point).

Typical Table Value

What is the typical table range for the majority of diamonds of ideal cut?

Answer: 53 to 57

What is the typical table range for the majority of diamonds of premium cut?

Answer: 58 to 62

Price vs. Clarity and Diamond Clarity

diamonds$volume <- with(diamonds, x * y * z)

ggplot(data = diamonds, aes(x = volume, y = price, color = clarity)) +
  geom_point() + 
  scale_color_brewer(type = 'div') + 
  scale_y_log10() + 
  scale_x_continuous(limits = c(0, quantile(diamonds$volume, 0.99)))
## Warning: Removed 540 rows containing missing values (geom_point).

Proportion of Friendships Initiated

pf$prop_initiated <- with(pf, friendships_initiated/friend_count)
summary(pf$prop_initiated)
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##  0.0000  0.4524  0.6250  0.6078  0.7838  1.0000    1962

prop_initiated vs. tenure

pf$year_joined <- with(pf, floor(2014 - (tenure/365)))
pf$year_joined.bucket <- cut(pf$year_joined, breaks = c(2004, 2009, 2011, 2012, 2014))
table(pf$year_joined.bucket)
## 
## (2004,2009] (2009,2011] (2011,2012] (2012,2014] 
##        6669       15308       33366       43658
ggplot(data = subset(pf, !is.na(tenure) & !is.na(prop_initiated)), aes(x = tenure, y = prop_initiated, color = year_joined.bucket)) + 
  geom_line(stat = 'summary', fun.y = median)

Smoothing prop_initiated vs. tenure

ggplot(data = subset(pf, !is.na(tenure) & !is.na(prop_initiated)), aes(x = tenure, y = prop_initiated, color = year_joined.bucket)) + 
  geom_line(stat = 'summary', fun.y = median, alpha = 1/5) + 
  geom_smooth()

Greatest prop_initiated Group

On average, which group initiated the greates proportion of its Facebook friendships?

Answer: People who joined after 2012. (The newest users)

Largest Group Mean prop_initiated

For the group with the largest proportion of friendships initiated, what is the group’s average proportion of friendships initiated?

by(pf$prop_initiated, pf$year_joined.bucket, summary)
## pf$year_joined.bucket: (2004,2009]
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##  0.0000  0.3418  0.4672  0.4668  0.5910  1.0000       5 
## -------------------------------------------------------- 
## pf$year_joined.bucket: (2009,2011]
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##  0.0000  0.3924  0.5357  0.5301  0.6750  1.0000      81 
## -------------------------------------------------------- 
## pf$year_joined.bucket: (2011,2012]
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##  0.0000  0.4576  0.6189  0.5985  0.7619  1.0000     408 
## -------------------------------------------------------- 
## pf$year_joined.bucket: (2012,2014]
##    Min. 1st Qu.  Median    Mean 3rd Qu.    Max.    NA's 
##  0.0000  0.5115  0.7018  0.6654  0.8490  1.0000    1468

Answer: 0.647

Why do you think this group’s proportion of friendships initiated is higher than the others?

Answer: They are new to Facebook so they are searching and adding all the people they know. People who have been on Facebook for a while have all of the new users add them. It is harder to find newer users than for new users to find older (tenure) users.

Price/Carat Binned, Faceted, & Colored

ggplot(data = diamonds, aes(x = cut, y = price/carat, color = color)) +
  geom_point(alpha = 1/10) + 
  geom_jitter() +
  facet_wrap(~clarity, ncol = 3) + 
  scale_color_brewer(type = 'div')