google_scholar_viz.Rmd

---
title: "google scholar viz"
output: html_notebook
---




```{r}

library(ggplot2)
library(dplyr)


```



```{r}

publications_google_scholar = readr::read_csv("_data/publications_google_scholar.csv")

publications_citations_per_years = readr::read_csv("_data/publications_citations_per_years.csv")



```

```{r export_citations}

affiliation_colors <- c(
  "Danone_Research" = "#11ACED", 
  "EMBL" = "#18974C", 
  "INRAE" = "#00a3a6", 
  "Institut_Pasteur" = "#005C97", 
  "UPEC" = "#D32F2F"
  
)


readr::read_csv2("_data/publications_affiliations.csv") %>%
  merge(publications_citations_per_years, by="pubid") %>%
  mutate(Year = as.character(Year)) %>%
  group_by(Year) %>%
  mutate(year_cite=sum(Citations)) %>%
  ggplot() + geom_bar(aes(x=Year, y=Citations, fill=primary_affiliation), stat="identity" ) +
  #geom_text(aes(x=Year, y=year_cite, label=year_cite), vjust=-0.5, color="black") + 
  scale_fill_manual("Primary affiliation", values = affiliation_colors) +
  ggtitle("Number of citations over time") +

  theme_classic() +
  theme(legend.position = c(0.02, 0.98),    # Top left position
        legend.justification = c(0, 1),     # Anchor point (top left corner)
        legend.background = element_rect(fill="transparent", color="transparent"), # Transparent background and border
        legend.margin = margin(t = 0, r = 0, b = 0, l = 0))  # Reduce margin to make it closer to the corner

ggsave("images/julien_tap_publications_citations_affiliations.png")

```


```{r eval=FALSE, include=FALSE}

# Load required libraries
library(tidyverse)

data <- publications_citations_per_years

# Define a function to compute h-index
compute_h_index <- function(citations) {
  sorted_citations <- sort(citations, decreasing = TRUE)
  h_index <- 0
  for (i in seq_along(sorted_citations)) {
    if (sorted_citations[i] >= i) {
      h_index <- i
    } else {
      break
    }
  }
  return(h_index)
}

# Compute h-index over time
years <- unique(data$Year)
h_indices <- numeric(length(years))

for (i in seq_along(years)) {
  year <- years[i]
  
  # Filter data for all years up to the current year
  subset_data <- data %>% filter(Year <= year)
  
  # Sum citations for each publication
  summed_citations <- subset_data %>% group_by(pubid) %>% summarize(total_citations = sum(Citations))
  
  # Compute h-index for the year
  h_indices[i] <- compute_h_index(summed_citations$total_citations)
}

# Combine results into a data frame
result <- data.frame(Year = years, H_index = h_indices)

result %>%
  arrange(Year) %>%
  #filter(Year != 2009) %>%
  mutate(Year= as.character(Year)) %>%
  ggplot() + geom_bar(aes(x=Year, y=H_index), stat="identity", fill="#0a0a0a") +
  geom_text(aes(x=Year, y=H_index, label=H_index), vjust=1.5, color="white") + # This line adds h-index values
  ylab("h-index") +
  theme_classic() +
  ggtitle("Hirsch h-index variations over time", subtitle = "*h* publications cited at least *h* times")




```

```{r eval=FALSE, include=FALSE}

citation_history = readr::read_csv("_data/citation_history.csv")

citation_history %>%
mutate(year = as.character(year)) %>%
  ggplot() + geom_bar(aes(x=year,y=cites), stat="identity", fill="#0a0a0a") +
  geom_text(aes(x=year, y=cites, label=cites), vjust=1.5, color="white") + # This line adds h-index values
  theme_classic()

#ggsave(filename = "images/julien_tap_citations_per_year.png")


```

```{r export_hindex, message=FALSE, warning=FALSE}
readr::read_csv2("_data/publications_affiliations.csv") %>%
  merge(publications_citations_per_years, by="pubid", all = TRUE) %>%
  select(pubid,primary_affiliation, Year, Citations) -> data

# Compute cumulative citations for each publication up to each year
cumulative_data <- data %>%
  group_by(pubid) %>%
  arrange(Year) %>%
  mutate(cumulative_citations = cumsum(Citations)) %>%
  ungroup()

# Compute h-index for each year based on cumulative citations
compute_h_index <- function(citations) {
  sorted_citations <- sort(citations, decreasing = TRUE)
  h_index <- sum(sorted_citations >= 1:length(sorted_citations))
  return(h_index)
}

h_index_data <- cumulative_data %>%
  group_by(Year) %>%
  mutate(h_index = compute_h_index(cumulative_citations),
         contributes_to_h_index = ifelse(cumulative_citations >= row_number(desc(cumulative_citations)) & row_number(desc(cumulative_citations)) <= h_index, 1, 0)) %>%
  ungroup()

# Aggregation by affiliation for publications that contributed to h-index
affiliation_contrib <- h_index_data %>%
  filter(contributes_to_h_index == 1) %>%
  group_by(Year, primary_affiliation) %>%
  summarise(count_contrib = n()) %>%
  ungroup()

affiliation_contrib %>%
  arrange(Year) %>%
  #filter(Year != 2009) %>%
  mutate(Year= as.character(Year)) %>%
  ggplot() + geom_bar(aes(x=Year, y=count_contrib, fill=primary_affiliation), stat="identity") +
  #geom_text(aes(x=Year, y=max_h_index, label=H_index), vjust=1.5, color="white") + # This line adds h-index values
  ylab("h-index") +
  theme_classic() +
  ggtitle("Hirsch h-index variations over time", subtitle = "*h* publications cited at least *h* times") +
  scale_fill_manual("Primary affiliation", values = affiliation_colors) +
  theme_classic() +
  theme(legend.position = c(0.02, 0.98),    # Top left position
        legend.justification = c(0, 1),     # Anchor point (top left corner)
        legend.background = element_rect(fill="transparent", color="transparent"), # Transparent background and border
        legend.margin = margin(t = 0, r = 0, b = 0, l = 0))  # Reduce margin to make it closer to the corner


ggsave("images/julien_tap_h_index_affiliations.png")

```