Statistics for poppr installations.

poppr_stats.r

if (!require('devtools')) install.packages('devtools'); require('devtools')
if (!require('data.table')) install.packages('data.table'); require('data.table')
install_github('installr', 'talgalili') # get the latest installr R package
# or run the code from here:
# https://github.com/talgalili/installr/blob/master/R/RStudio_CRAN_data.r

require(installr)
if(!require(ISOcodes)) install.packages("ISOcodes", repos = "http://cran.at.r-project.org"); require(ISOcodes)

folder <- "~/Documents/poppr_stats"
current_files <- dir(folder)
poppr_data <- list()

if(length(current_files) == 0 | file.info(current_files[1])$size == 0){ 
  start_date <- '2013-04-26'
} else {
  start_date <- strsplit(current_files[length(current_files)], "[.]")[[1]][1]
  start_date <- substr(start_date, 12, nchar(start_date))
  poppr_data[["old"]] <- read.table(current_files[length(current_files)], header = TRUE, 
                                sep = ",", quote = "\"", dec = ".", fill = TRUE, 
                                stringsAsFactors = FALSE, comment.char = "", 
                                as.is=TRUE, check.names = FALSE)
}
print(start_date)
if(start_date != Sys.Date()){
  folder <- download_RStudio_CRAN_data(START = start_date, END = Sys.Date() - 1, log_folder = folder)
}

out_file <- paste("poppr_data_", Sys.Date(), ".csv", sep = "")

if (start_date != Sys.Date()){
  poppr_data[["new"]] <- read_RStudio_CRAN_data(folder)
  my_RStudio_CRAN_data <- rbindlist(poppr_data)
} else {
  my_RStudio_CRAN_data <- poppr_data[["old"]]
}

barplot_package_users_per_day("poppr", my_RStudio_CRAN_data)

poppr_data <- my_RStudio_CRAN_data[my_RStudio_CRAN_data$package %in% "poppr", ]
unlink("*gz")
write.table(poppr_data, file = out_file, sep = ",", row.names=FALSE, col.names=TRUE)
unique_poppr <- poppr_data[!duplicated(poppr_data$ip_id), ]

data(ISO_3166_1)
isoidx <- which(ISO_3166_1$Alpha_2 %in% unique_poppr$country)
iso <- ISO_3166_1[isoidx, ]


invisible(lapply(iso$Alpha_2, function(x) 
	                             unique_poppr$country[which(unique_poppr$country == x)] <<- iso[iso$Alpha_2 == x, "Name"]))


unique_poppr$total <- rep(cumsum(table(unique_poppr$date)), table(unique_poppr$date))
require(ggplot2)

by_date <- ggplot(unique_poppr) + geom_histogram(aes(x = date, fill = country)) + theme_bw() + theme(axis.text.x=element_text(size = 10, angle=45, hjust=1, vjust=1)) + labs(title = paste("Poppr installations by date as of", Sys.Date(), sep = "\n")) + annotate("text", x = 1, y = 6, label = paste("Total:", nrow(unique_poppr)), hjust = 0, vjust = 1.1, size = 12)
res <- paste(folder,"results", sep = "_")

ggsave(plot = by_date, file = paste(res, "/inst_by_date_", Sys.Date(), ".pdf", sep = ""), scale = 2, width = 11, height = 8.5)

by_country <- ggplot(unique_poppr) + geom_histogram(aes(x = country, fill = country)) + theme_bw() + theme(axis.text.x=element_text(size = 10, angle=45, hjust=1, vjust=1)) + labs(title = paste("Poppr installations by country as of", Sys.Date(), sep = "\n")) + annotate("text", x = 1, y = 10, label = paste("Total:", nrow(unique_poppr)), hjust = 0, vjust = 1.1, size = 12)

ggsave(plot = by_country, file = paste(res, "/inst_by_country_", Sys.Date(), ".pdf", sep = ""), scale = 2, width = 11, height = 8.5)

by_total <- ggplot(unique_poppr) + geom_point(aes(x = date, y = total, color = version)) + theme_bw() + theme(axis.text.x=element_text(size = 10, angle=45, hjust=1, vjust=1)) + labs(title = paste("Total Poppr installations as of", Sys.Date(), sep = "\n")) + annotate("text", x = 1, y = nrow(unique_poppr), label = paste("Total:", nrow(unique_poppr)), hjust = 0, vjust = 1.1, size = 12, color = "red")

ggsave(plot = by_total, file = paste(res, "/inst_total_", Sys.Date(), ".pdf", sep = ""), scale = 2, width = 11, height = 8.5)