data_analysis_rq2.Rmd

---
title: "RQ2: How are the active versions distributed among the different releases of a library?"
output: html_notebook
---



```{r}
library(tidyverse)
library(latex2exp)
library(scales)
theme_set(theme_bw())
```

### Distribution of active, passive and dormant versions

```{r}
# df_patterns <- metrics %>%
#   filter(is.na(isPassive) == F) %>% 
#   mutate(Library = paste(groupId, artifactId, sep = ":")) %>%
#   mutate(PA = ifelse(isPassive == "true", "P", "A")) %>%
#   group_by(Library) %>%
#   arrange(releaseDate) %>%
#   summarise(PAList = list(PA), dates = list(releaseDate))

# position of the active versions
PAList <- df_patterns$PAList
PosA <- c()
PosP <- c()
for (i in 1:length(PAList)) {
  PALib <- PAList[i][[1]]
  for (j in 1:length(PALib)) {
    if (PALib[j] == "A") {
      PosA <- c(PosA, j / length(PALib))
    }
    # else{#is passive
    #   PosP <- c(PosP, j / length(PALib))
    # }
  }
}

a <- character(0)
patterns_file <- patterns_file %>% filter(Patterns != "")
PAList <- patterns_file$newlistPass
PosA <- c()
PosP <- c()
for (i in 1:length(PAList)) {
   print(i)
  PALib <- PAList[i][[1]]
  for (j in 1:length(PALib)) {
    if (PALib[j] == "A") {
      PosA <- c(PosA, j / length(PALib))
    }
    # else{#is passive
    # PosP <- c(PosP, j / length(PALib))
    # }
  }
}
  
data.frame(PosA = PosA) %>% 
  ggplot(aes(PosA)) +
  geom_histogram() +
geom_histogram(bins = 30, color="black", fill = "#CCCCCC") +
   ylab("count") +
   xlab("relative position of active versions")

ggsave(filename = "Figures/hist_pos_active_versions.pdf", 
       height = 2.5, width = 4,  units = c("in"), device = "pdf")

data.frame(PosA = PosA) %>% 
  filter(PosA == 1)


write_csv(passive[1:387,], "passive_non_dormant_libs.csv")

```


### Patterns search

```{r}

patterns_file <- read_csv("Data/patterns.csv")
patterns_file <- patterns_file %>% select(library, listPass) %>% rename(Library = library)
patterns_file <- inner_join(patterns_file, libraries_betw5and200_versions, by = "Library") 
patterns_file$newlistPass <- substring(patterns_file$listPass, 2, nchar(patterns_file$listPass)-1)
patterns_file$newlistPass <- gsub("true", "P", patterns_file$newlistPass)
patterns_file$newlistPass <- gsub("false", "A", patterns_file$newlistPass)
patterns_file$newlistPass <- strsplit(patterns_file$newlistPass, ",")
# get the patterns
patterns <- c()
for (i in 1:length(patterns_file$newlistPass)) {
  patterns <- c(patterns, paste(rle(patterns_file$newlistPass[i][[1]])$values, collapse = ""))
}
patterns_file$Patterns <- patterns
patterns_file$Patterns <- as.factor(patterns_file$Patterns)
table(patterns_file$Patterns)
count(patterns_file, Patterns) %>% arrange(desc(n))


View(patterns_file %>% filter(Patterns == "APA") %>% select(Library, MeanPageranks, Patterns))

xxx <- read_csv("Data/patterns.csv")
View(xxx %>% filter(library == "com.google.guava:guava-jdk5"))
```