RCran.R

install.packages("rvest")
install.packages("xml2")
install.packages("installr")
install.packages("ggplot2")
install.packages("data.table")
install.packages("plyr")
install.packages("cranlogs")
install.packages("lubridate")
install.packages("magrittr")
install.packages("NLP")
install.packages("tm")
install.packages("wordcloud")
install.packages("RColorBrewer")
install.packages("SnowballC")
install.packages("miniCRAN")
install.packages("igraph")
install.packages("cowplot")

library(rvest)
library(xml2)
library(installr)
library(ggplot2)
library(data.table)
library(plyr)
library(cranlogs)
library(lubridate)
library(magrittr)
library(NLP)
library(tm)
library(wordcloud)
library(RColorBrewer)
library(SnowballC)
library(miniCRAN)
library(igraph)
library(cowplot)

# Names of packages AND respective summary available on https://cran.r-project.org/
CRAN_package_wbpg <- read_html("https://cran.r-project.org/web/packages/available_packages_by_name.html")
  
# storing and cleaning package names
package_names <- CRAN_package_wbpg %>%
  html_nodes("a") %>%
  html_text()
package_names <- package_names[!package_names %in% LETTERS]

# storing and cleaning package summary
package_summary <- CRAN_package_wbpg %>%
  html_nodes("td+ td") %>%
  html_text()
package_summary <- gsub("[\r\n]", "", package_summary)
  
# creating a data frame 
CRAN_package_data <- data.frame(
  Name = package_names, Summary = package_summary)

###################################################################################################################################

# set your working directory and package stats file will be stored in folder "Logs"
dir.create("Logs")
RStudio_CRAN_dir <- download_RStudio_CRAN_data(START = '2020-01-01',END = '2020-01-10', log_folder="Logs")

RStudio_CRAN_data <- read_RStudio_CRAN_data(RStudio_CRAN_dir)

dim(RStudio_CRAN_data)

# creating most downloaded package list
pkg_list <- most_downloaded_packages(RStudio_CRAN_data)
pkg_list

lineplot_package_downloads(names(pkg_list),RStudio_CRAN_data)

# bar plot for day to day package download stats comparison for "ggplot" and "Rcpp"
par(mfrow=c(1,2))
barplot_package_users_per_day("ggplot",RStudio_CRAN_data)
barplot_package_users_per_day("Rcpp", RStudio_CRAN_data)

###################################################################################################################################

# assigned a certain time period
total_downld <- cran_downloads(from = "2019-01-01", to = "2019-12-31")
# setting mode to data.table
setDT(total_downld)
# converting the date to week and month format
total_downld[, `:=`(
  round_week = floor_date(date, "week"),
  round_month = floor_date(date, "month")
)]

# total downloads
total_downld[, .(total =sum(count))]

# creating a random color generator for graph plot
random_col <- function(n){
  sample(viridis::viridis(100), n)
}

# Weekly download analysis 
total_downld[, .(count = sum(count)), round_week] %>%
  ggplot(aes(round_week, count)) + 
  geom_col(fill = random_col(1)) + 
  labs(
    title = "Packages downloads by Week on RStudio CRAN mirror", 
    subtitle = "data via {cranlogs}", 
    x = "Week classification", y="Number of downloads"
  ) + 
  theme_minimal()

# Monthly download analysis 
total_downld[, .(count = sum(count)), round_month] %>%
  ggplot(aes(round_month, count)) + 
  geom_col(fill = random_col(1)) + 
  labs(
    title = "Packages downloads by Month on RStudio CRAN mirror", 
    subtitle = "data via {cranlogs}", 
    x = "Month classification", y="Number of downloads"
  ) + 
  theme_minimal()

###################################################################################################################################
# R DOWNLOADS

# assigned a certain time period
total_R <- cran_downloads("R", from = "2019-01-01", to = "2019-12-31")
# setting mode to data.table
setDT(total_R)
# converting the date to week and month format
total_R[, `:=`(
  round_week_r = floor_date(date, "week" ),
  round_month_r = floor_date(date, "month" )
) ]

# total downloads
total_R[, .(total = sum(count))]

# Weekly download analysis 
total_R[, .(count = sum(count)), round_week_r] %>%
  ggplot(aes(round_week_r, count)) + 
  geom_col(fill = random_col(1)) + 
  labs(
    title = "R downloads by Week on RStudio CRAN mirror", 
    subtitle = "data via {cranlogs}", 
    x = "Week classification", y="Number of downloads"
  ) + 
  theme_minimal()

# Monthly download analysis 
total_R[, .(count = sum(count)), round_month_r] %>%
  ggplot(aes(round_month_r, count)) + 
  geom_col(fill = random_col(1)) + 
  labs(
    title = "R downloads by Month on RStudio CRAN mirror", 
    subtitle = "data via {cranlogs}", 
    x = "Month classification", y="Number of downloads"
  ) + 
  theme_minimal()

###################################################################################################################################

# r version downloads

# R version plot analysis
total_R[, .(count = sum(count)), version][order(count, decreasing = TRUE)] %>%
  head(10) %>% 
  ggplot(aes(reorder(version, count), count)) +
  coord_flip() +
  geom_col(fill = random_col(1)) + 
  labs(
    title = "10 most downloaded R versions in 2019 on RStudio CRAN mirror", 
    subtitle = "data via {cranlogs}", 
    x = "version", y="Number of downloads"
  ) + 
  theme_minimal()

###################################################################################################################################

# word-cloud of package keywords

# converting data to string
data <- toString(package_summary)
data <- gsub('[[:punct:] ]+',' ',data)

# cleaning data
docs <- VCorpus(VectorSource(data))
docs <- tm_map(docs, content_transformer(tolower))
docs <- tm_map(docs, removeNumbers)
docs <- tm_map(docs, removeWords, stopwords("english"))
docs <- tm_map(docs, removePunctuation)
docs <- tm_map(docs, stripWhitespace)
docs <- tm_map(docs, stemDocument)
docs <- tm_map(docs, PlainTextDocument)

# creating a matrix for tabulation of word with respective frequency 
dtm <- TermDocumentMatrix(docs)
m <- as.matrix(dtm)
v <- sort(rowSums(m),decreasing=TRUE)
d <- data.frame(word = names(v),freq=v)
head(d, 10)

# plotiing word-cloud
set.seed(1234)
wordcloud(words = d$word, freq = d$freq, min.freq = 1,
          max.words=550, random.order=FALSE, rot.per=0.35, 
          colors=rev(colorRampPalette(brewer.pal(9,"Blues"))(32)[seq(8,32,6)]))

###################################################################################################################################

# Network-graph of package dependencies

# retrieving the available packages on CRAN
pkgdata <- pkgAvail(repos = c(CRAN="http://cran.revolutionanalytics.com"), 
                    type="source")
head(pkgdata[, c("Depends", "Suggests")])

# packages to check dependencies on
tags <- c("data.table", "chron", "plyr")

# plotting network graph
set.seed(50)
plot(makeDepGraph(tags, includeBasePkgs=FALSE, suggests=TRUE, enhances=TRUE), 
     legendPosEdge = c(-1, 1), legendPosVertex = c(1, 1), vertex.size=9)

###################################################################################################################################

# Most popular package - last-day, last-week, last-month

# storing number of download's for packages in last-day, last-week, last-month
mp_pkg_day <- cran_top_downloads(when = "last-day", count = 10)
mp_pkg_week <- cran_top_downloads(when = "last-week", count = 10)
mp_pkg_month <- cran_top_downloads(when = "last-month", count = 10)

# last-day plot
day_plot <- ggplot(data=mp_pkg_day, aes(x=package, y=count)) +
  geom_bar(stat="identity", fill="steelblue")+
  geom_text(aes(label=count), vjust=1.6, color="white", size=3.5)+
  theme_minimal()+
  theme(axis.text.x=element_text(angle=45, hjust=1))+
  labs(title = "Most popular packages downloaded on last day",
        x = "Packages", y = "Number of downloads")

# last-week plot
week_plot <- ggplot(data=mp_pkg_week, aes(x=package, y=count)) +
  geom_bar(stat="identity", fill="#999999")+
  geom_text(aes(label=count), vjust=1.6, color="white", size=3.5)+
  theme_minimal()+
  theme(axis.text.x=element_text(angle=45, hjust=1))+
  labs(title = "Most popular packages downloaded on last week",
       x = "Packages", y = "Number of downloads")

# last-month plot
month_plot <- ggplot(data=mp_pkg_month, aes(x=package, y=count)) +
  geom_bar(stat="identity", fill="#E69F00")+
  geom_text(aes(label=count), vjust=1.6, color="white", size=3.5)+
  theme_minimal()+
  theme(axis.text.x=element_text(angle=45, hjust=1))+
  labs(title = "Most popular packages downloaded on last month",
       x = "Packages", y = "Number of downloads")

# plotting graph together using library "cowplot" 
plot_grid(day_plot, week_plot, month_plot, labels = "AUTO")