Amount of FWO-SB projects per Flemish university

FWO-SB is a very important funding channel for PhD students in Flanders (Belgium). After submitting a proposal in September, PhD candidates have to defend their project for a jury in November/December. Usually the results appear online in mid-December, and for this year’s call that was today:

De resultaten SB beurzen oproep 2017 staan online https://t.co/2EG16ohL9I

— FWO (@FWOVlaanderen) December 14, 2017

Since I was a little bit curious about how many projects went to each university and since I want to keep my tidyverse skills up to date, I parsed the PDF-file with all of the results and plotted it using  ggplot2.

Good job KU Leuven!

UPDATE

Pieter Meysman asked me to normalise the data a little bit, so I looked for the amount of professors per university and plotted the data again. This time, the x-axis represents the amount of projects per 100 professors:

I’ve used the following amount of professors (click  on link for source):

• UAntwerpen: 624
• KU Leuven: 1561
• UGent: 1502
• VUB: 402
• UHasselt: 392

The code to create this plot can be found here and on Github:

 

	library(tidyverse)
	library(pdftools)
	df <- pdf_text("SB-Beurs-2017.pdf") %>%
	str_split(pattern = "\n") %>%
	unlist() %>%
	as_tibble() %>%
	filter(str_detect(value, "18N")) %>% # select lines with project number in them
	mutate(value = str_trim(value), # trim trailing spaces
	value = if_else(str_sub(value, 1, 2) %in% str_c(LETTERS, " "), # Remove letters
	str_sub(value, 3, -1),
	value)) %>%
	separate(value, into = c("Mandaathouder", "Universiteit", "Aanvraagnummer"), fill = "right", sep = "\\s{2,}") %>%
	mutate(Aanvraagnummer = if_else(is.na(Aanvraagnummer), # Not everything is parsed well so neet some tweaking
	str_sub(Universiteit, -8,-1),
	Aanvraagnummer),
	Universiteit = if_else(str_detect(Universiteit, "18N"), # Split strings with project numbers in their University column
	str_sub(Universiteit, 1, -9),
	Universiteit),
	Universiteit = if_else(Universiteit == "", # Split strings with University name in Mandaathouder column
	str_extract(Mandaathouder, "[A-z]+$"),
	Universiteit),
	Universiteit = if_else(Universiteit == "Leuven", # Fix if only Leuven was put in the University column
	"KU Leuven",
	str_trim(Universiteit)))
	df %>% group_by(Universiteit) %>%
	summarise(Total_projects = n()) %>%
	ggplot(aes(x = reorder(Universiteit, Total_projects), y = Total_projects)) +
	geom_col() +
	xlab("") +
	ylab("Amount of FWO-SB projects") +
	coord_flip() +
	theme_minimal() +
	theme(axis.text.y = element_text(size = 16, face = "bold"))
	ggsave("FWO_SB_2017.png")
	# Update
	professors <- tibble(Universiteit = c("UAntwerpen", "KU Leuven", "UGent", "VUB", "UHasselt"),
	NumberZAP = c(624, 1561, 1502, 402, 392))
	df %>% group_by(Universiteit) %>%
	summarise(Total_projects = n()) %>%
	ungroup() %>%
	left_join(professors) %>%
	mutate(normalized_count = Total_projects/(NumberZAP/100)) %>%
	ggplot(aes(x = reorder(Universiteit, normalized_count), y = normalized_count)) +
	geom_col() +
	xlab("") +
	ylab("Amount of FWO-SB projects per 100 professors") +
	coord_flip() +
	theme_minimal() +
	theme(axis.text.y = element_text(size = 16, face = "bold"))
	ggsave("FWO_SB_2017_per100prof.png")

view raw

FWO_SB_2017.R

hosted with ❤ by GitHub