This notebook is part of a GitHub repository: https://github.com/pessini/SFI-Grants-and-Awards
MIT Licensed
Author: Leandro Pessini
SFI - Awards Distribution Analysis
¶An overview of science funding distribution in Ireland¶
This report will provide an overview of awards applied by Science Foundation Ireland (SFI) which is the national foundation for investment in scientific and engineering research. The data provided covers a period of time from 2000 to 2019.
The Agreed Programme for Government, published June 2002, provided for establishing SFI as a separate legal entity. In July 2003, SFI was established on a statutory basis under the Industrial Development (Science Foundation Ireland) Act, 2003.
SFI provides awards to support scientists and engineers working in the fields of science and engineering that underpin biotechnology, information and communications technology and sustainable energy and energy-efficient technologies.
The focus of this report is on research funding and geographical distribution of grant awardees.
Audience¶
A core principle of data analysis is understanding your audience before designing your visualization. It is important to match your visualization to your viewer’s information needs.
This ad hoc analysis aims to deliver a presentation to the SFI Board Members. The Board has several responsibilities which include the revision of strategies of the Agency and major plans of action. One of their major functions is to establish the Agency's direction and how the resources are allocated.
The dataset used is Science Foundation Ireland Grant Commitments and it details all STEM (science, technology, engineering and maths) research projects funded by Science Foundation Ireland (SFI) since its foundation in 2000. For more information, check out the Data Dictionary available.
Dataset provided by Ireland's Open Data Portal which helds public data from Irish Public Sectors such as Agriculture, Economy, Housing, Transportation etc.
Libraries¶
# Change the default plots size
options(repr.plot.width=15, repr.plot.height=10)
options(warn=-1)
# Suppress summarise info
options(dplyr.summarise.inform = FALSE)
options(dplyr = FALSE)
# Check if the packages that we need are installed
want = c("dplyr", "ggplot2", "ggthemes", "gghighlight",
"grid", "foreign", "scales", "ggpubr", "forcats",
"stringr", "lubridate")
have = want %in% rownames(installed.packages())
# Install the packages that we miss
if ( any(!have) ) { install.packages( want[!have] ) }
# Load the packages
junk <- lapply(want, library, character.only = T)
# Remove the objects we created
rm(have, want, junk)
# Importing dataset
sfi.grants <- read.csv('../data/Open-Data-Final.csv')
# Checking dataset's structure
head(sfi.grants)
# Cleaning data
sfi.new.grants <- sfi.grants
# Because we are going to analyze new grants from the SFI...
# There are a few grants from funding partners and a few splitting of awards where an award
# was transferred from one institution to another.
# Removing those
sfi.new.grants <- sfi.new.grants[-grep("(N)", sfi.grants$Proposal.ID, fixed = T),]
sfi.new.grants <- sfi.new.grants[-grep("(X)", sfi.grants$Proposal.ID, fixed = T),]
sfi.new.grants <- sfi.new.grants[-grep("(T)", sfi.grants$Proposal.ID, fixed = T),]
# Total of rows from Partners Funds and Splittings - 68
nrow(sfi.grants) - nrow(sfi.new.grants)
# There are 11 "negatives" or 0 grants which will interfere in the analysis
# Decided to remove those (only one will be transform to absolute value)
negatives <- sfi.new.grants %>% filter(Revised.Total.Commitment < 10)
# Remove all negative grants
sfi.new.grants$Revised.Total.Commitment <- abs(sfi.new.grants$Revised.Total.Commitment)
sfi.new.grants <- sfi.new.grants[sfi.new.grants$Revised.Total.Commitment > 10,]
sfi.new.grants <- sfi.new.grants %>% mutate(Programme.Name.Clean = str_replace(Programme.Name, "SFI ", ""),
Programme.Name.Clean = str_replace(Programme.Name.Clean, " Programme", ""))
sfi.new.grants$Date <- as.Date(sfi.new.grants$Start.Date, format = "%d/%m/%Y")
sfi.new.grants$Date <- format(sfi.new.grants$Date, "%Y-%m-%d")
allMissing <- is.na(sfi.new.grants)
#get a count for each column
counts <- colSums(allMissing)
counts
paste0("Number of rows in the dataset: ", nrow(sfi.new.grants))
sfi.new.grants %>% mutate(year = year(Date)) %>%
group_by(year) %>%
summarise(total = sum(Revised.Total.Commitment),
n= n(), average = sum(total)/sum(n)) %>%
ggplot( aes(x = year, y= total) ) +
geom_line( color="#69b3a2", size = 1) +
geom_point( color="#69b3a2",size=3) +
labs(x = "Year", y = "",
title = "Total amount awarded throughout the years",
subtitle = "Amount awarded (€)") +
scale_y_continuous(labels = scales::label_number_si(accuracy=0.1)) +
scale_x_continuous(breaks = seq(from = 2000, to = 2019, by = 1)) +
scale_fill_brewer(palette='Dark2') +
theme_minimal() +
theme(axis.text.x = element_text(face="bold", color="#636363", size=12),
axis.text.y = element_text(face="bold", color="#636363", size=12),
plot.title=element_text(vjust=1.5, family='', face='bold', colour='#636363', size=20),
plot.subtitle=element_text(vjust=1.5, family='', face='bold', colour='#636363', size=15))
Most awarded Institutes by amount ۦ
# Grouping by Institutes and creating the sum, mean and total number of grants
by_institute <- sfi.new.grants %>%
group_by(Research.Body) %>%
summarise(total = sum(Revised.Total.Commitment),
mean = mean(Revised.Total.Commitment),
n= n()) %>% ungroup()
# Plot the Top 10 Institutes which received the highest grants amount
top10.institutes.value <- by_institute %>% arrange(desc(total)) %>% head(10)
top10.institutes.value %>%
ggplot( aes(x =reorder(as.factor(Research.Body), total),
y= total, fill="") ) +
geom_bar(inherit.aes = TRUE, lineend = 'round',
stat = "identity", width = .5, alpha=.9) +
scale_y_continuous(labels = scales::label_number_si(accuracy=0.1)) +
annotate("segment", x = 8, xend = 8, y = 437562222, yend = 749722478,
arrow = arrow(ends = "both", angle = 90, length = unit(.5,"cm"))) +
annotate("curve", curvature = -.3, x = 9.7, xend = 8.1, y = 6e+08, yend = 6e+08,
colour = "#636363", size = 2, arrow = arrow()) +
annotate("text", x = 7, y = 6e+08, family = "", fontface = 3, size=6,
label = "Huge gap between \n the most awarded institutions\n Dublin x Outside-Dublin") +
scale_fill_brewer(palette='Dark2') +
labs(x = "", y = "",
title = "Top 10 Granted Institutes - Amount awarded (€)",
subtitle = "The grants values are in Millions") +
theme_minimal() + coord_flip() +
theme(legend.position = "none",
axis.text.x = element_text(face="bold", color="#636363", size=12),
axis.text.y = element_text(face="bold", color="#636363", size=12),
plot.title=element_text(vjust=1.5, family='', face='bold', colour='#636363', size=25),
plot.subtitle=element_text(vjust=1.5, family='', face='bold', colour='#636363', size=15))
From the chart we can see that Trinity College Dublin was awarded almost the double amount comparing to the first top College outside Dublin (University College Cork).
Most awarded Institutes by number of grants¶
top10.institutes.total <- by_institute %>% arrange(desc(total)) %>% top_n(10)
top10.institutes.total %>%
ggplot( aes(x =reorder(as.factor(Research.Body), n),
y= n) ) +
geom_bar(lineend = 'round',
stat = "identity", width = .5, alpha=.9, fill="steelblue") +
annotate("segment", x = 8, xend = 8, y = 601, yend = 1150,
arrow = arrow(ends = "both", angle = 90, length = unit(.5,"cm"))) +
annotate("curve", curvature = -.3, x = 9.7, xend = 8.1, y = 1100, yend = 1100,
colour = "#636363", size = 2, arrow = arrow()) +
annotate("text", x = 7, y = 900, family = "", fontface = 3, size=6,
label = "Almost 2x gap between \n the most awarded institutions\n Dublin vs Outside-Dublin") +
scale_fill_brewer(palette='Set2') +
scale_y_continuous(name="Number of Grants") +
labs(x = "", y = "",
title = "Top 10 Granted Institutes - Number of Grants",
subtitle = "") +
theme_minimal() + coord_flip() +
theme(legend.position = "none",
axis.text.x = element_text(face="bold", color="#636363", size=12),
axis.text.y = element_text(face="bold", color="#636363", size=12),
plot.title=element_text(vjust=1.5, family='', face='bold', colour='#636363', size=25),
plot.subtitle=element_text(vjust=1.5, family='', face='bold', colour='#636363', size=15))
Comparing number of Grants awarded, Trinity College is again the winner. Has received almost double of grants in numbers comparing to University College Cork, which is the first top outside-Dublin institution.
Grants distribution by Programmes¶
# Grouping by Programmes and creating the sum, mean and total number of grants
by_programme <- sfi.new.grants %>%
group_by(Programme.Name.Clean) %>%
summarise(total = sum(Revised.Total.Commitment),
mean = mean(Revised.Total.Commitment),
n= n()) %>% ungroup()
top10.programmes.value <- by_programme %>% arrange(desc(total)) %>% head(10)
top10.programmes.value %>%
ggplot( aes(x =reorder(as.factor(Programme.Name.Clean), total),
y= total, fill="") ) +
geom_bar(lineend = 'round',
stat = "identity", width = .5,
alpha= ifelse(top10.programmes.value$Programme.Name.Clean == "Principal Investigator" | top10.programmes.value$Programme.Name.Clean == "Research Centres",
.9, .4)) +
scale_fill_brewer(palette='Dark2') +
scale_y_continuous(labels = scales::label_number_si(accuracy=0.1)) +
annotate("text", x = 8, y = 4.7e+08, family = "", fontface = 3, size=6,
label = "Average amount awarded \n Principal Investigator = 1.1M \n Research Centres = 7.5M") +
labs(x = "", y = "",
title = "Leading programmes are Principal Investigator and Research Centres",
subtitle = "Total amount awarded (€)") +
theme_minimal() + coord_flip() +
theme(legend.position = "none",
axis.text.x = element_text(face="bold", color="#636363", size=12),
axis.text.y = element_text(face="bold", color="#636363", size=12),
plot.title=element_text(vjust=1.5, family='', face='bold', colour='#636363', size=20),
plot.subtitle=element_text(vjust=1.5, family='', face='bold', colour='#636363', size=15))
The 2 most granted Programmes are: SFI Principal Investigator Programme and SFI Research Centres Programme.
Although SFI Principal Investigator Programme has the highest total amount awarded, the average amount awarded to SFI Research Centres Programme is 7x higher than the SFI Principal Investigator Programme.
This Dashboard was created using Tableau® software.
The focus is to show the main insights found on this analysis. The Dashboard along with the Data Exploration can be found on Tableau website.
GitHub repository
Author: Leandro Pessini
R.version$version.string