index.Rmd

---
title: "Open Science MOOC"
output:
  flexdashboard::flex_dashboard:
    orientation: rows
    css: style.css
    logo: mooc-logo.png
    favicon: mooc-logo.png
    navbar:
      - { title: "About", href: "https://dataplanes.org" }
    source_code: https://github.com/lhehnke/osmooc-dashboard

---

```{r, include = FALSE}
# Load flexdashboard
library(flexdashboard)

# Install and load packages using pacman
if (!require("pacman")) install.packages("pacman")
library(pacman)

p_load(ggraph, igraph, kableExtra, networkD3, plotly, tidygraph, tidyverse)

# Set directory
MAIN_DIR <- rprojroot::find_rstudio_root_file()

# Import data
load(file = paste(MAIN_DIR, "osmooc-github.RData", sep = "/"))
```

Sidebar {.sidebar}
=======================================================================

### Hello, Open Scientist!

This dashboard gives you an overview of the repository statistics and user activities of the **[Open Science MOOC](https://opensciencemooc.eu/)** on [GitHub](https://github.com/opensciencemooc).

If you want to become part of *#TeamOpen*, you can join our ever-growing [Slack community](https://openmooc-ers.slack.com/) or follow us on [Twitter](https://twitter.com/opensciencemooc) for daily news on all things Open Science.

*Data for the analysis was collected on June 07, 2019.*


Overview {data-icon="fa-chart-line"}
=======================================================================

Row
-----------------------------------------------------------------------

### Module repositories {.value-box}

```{r}
repos <- nrow(repos_df_mod)
valueBox(repos, icon = "fa-github", color = "#91aac3")
```

### Stars {.value-box}

```{r}
stars <- repos_df_mod %>%
  summarize(sum(stargazers_count, na.rm = TRUE))
valueBox(stars, icon = "fa-star", color = "#91aac3")
```

### Forks {.value-box}

```{r}
forks <- repos_df_mod %>%
  summarize(sum(forks_count, na.rm = TRUE))
valueBox(forks, icon = "fa-code-fork", color = "#91aac3")
```

Row {.tabset .tabset-fade data-height=350}
-----------------------------------------------------------------------

### Which modules are starred the most?

```{r}
# Plot most starred repos
p <- repos_df_mod %>%
  select(name, stargazers_count) %>%
  group_by(name) %>%
  rename(Repository = name, Stars = stargazers_count) %>%
  ungroup() %>%
  mutate(Repository = reorder(Repository, Stars)) %>%
  ggplot(aes(Repository, Stars)) +
  geom_bar(stat = "identity", width = 0.5, alpha = 0.8, color = "#2c3e50", fill = "#2c3e50") +
  xlab("") + ylab("") +
  coord_flip() + theme_minimal()

ggplotly(p)
```

### Which modules are forked the most?

```{r}
# Plot most forked repos
p2 <- repos_df_mod %>%
  select(name, forks_count) %>%
  group_by(name) %>%
  rename(Repository = name, Forks = forks_count) %>%
  ungroup() %>%
  mutate(Repository = reorder(Repository, Forks)) %>%
  ggplot(aes(Repository, Forks)) +
  geom_bar(stat = "identity", width = 0.5, alpha = 0.8, color = "#2c3e50", fill = "#2c3e50") +
  xlab("") + ylab("") +
  coord_flip() + theme_minimal()

ggplotly(p2)
```

Contributors | Collaboration network {data-icon="fa-users"}
=======================================================================

Row
-----------------------------------------------------------------------

### Contributions {.value-box}

```{r}
contributions <- contributors_df %>%
  select(contributions) %>%
  summarize(sum(contributions, na.rm = TRUE)) %>%
  pull()
valueBox(contributions, icon = "fa-tasks", color = "#91aac3")
```

### Contributors {.value-box}

```{r}
collaborators <- contributors_df %>%
  select(login) %>%
  unique() %>%
  nrow()
valueBox(collaborators, icon = "fa-user", color = "#91aac3")
```

### Collaborations {.value-box}

```{r}
collaborations <- contributors_df %>%
  select(df_id, login) %>%
  rename(repository = df_id, name = login) %>%
  widyr::pairwise_count(name, repository, sort = TRUE, upper = FALSE) %>%
  nrow()
valueBox(collaborations, icon = "fa-users", color = "#91aac3")
```

Row {data-height=430}
-----------------------------------------------------------------------

### What does the collaboration network look like? {data-width=700}

This interactive graph shows the GitHub collaboration network. The top three contributors who collaborated the most are highlighted. Hover over a node to see the collaborator's name.

***

```{r, include = FALSE}
# Build edgelist
collab <- contributors_df %>%
  select(df_id, login) %>%
  rename(repository = df_id, name = login) %>%
  widyr::pairwise_count(name, repository, sort = TRUE, upper = FALSE)

# Convert to graph object
collab_graph <- graph_from_data_frame(collab, directed = FALSE) %>%
  as_tbl_graph()
```

```{r, include = FALSE}
# Compute network metrics
#members <- membership(cluster_walktrap(collab_graph))
degree <- igraph::degree(collab_graph)
group <- ifelse(degree >= 40, "top", "regular")
weight <- E(collab_graph)$n

# Convert igraph graph to networkD3 object
collab_d3 <- igraph_to_networkD3(collab_graph, group = group)

# Set metrics as attributes
collab_d3$nodes$degree <- degree
collab_d3$links$weight <- weight

# Set custom color range for grouping variable
colors <- JS('d3.scaleOrdinal(["#7d3945", "#2c3e50"])')
```

```{r}
# Plot interactive network
collab_network_d3 <- forceNetwork(Links = collab_d3$links,
                                  Nodes = collab_d3$nodes,
                                  Source = "source",
                                  Target = "target",
                                  NodeID = "name",
                                  Nodesize = "degree",
                                  Group = "group",
                                  colourScale = colors, # node colors based on "group"
                                  opacity = 0.8,
                                  opacityNoHover = 0.1,
                                  fontSize = 20,
                                  fontFamily = "Lato",
                                  Value = "weight",
                                  linkColour = "#91aac3",
                                  linkDistance = 150,
                                  linkWidth = JS("function(d) { return d.value*2; }"),
                                  charge = -50,
                                  zoom = TRUE)
collab_network_d3
```

### Which contributors collaborated the most? {data-width=340}

```{r}
collab_graph %>%
  mutate(centrality = centrality_degree(weight = n)) %>%
  as_tibble() %>%
  arrange(-centrality, name) %>%
  top_n(10, centrality) %>%
  rename(Contributor = name, "Degree centrality" = centrality) %>%
  kable(escape = FALSE) %>%
  kable_styling(bootstrap_options = c("striped", "hover"))
```{r}