MATH2270 Assignment 2
Visualising Open Data
Student Details
- Alistair Grevis-James (s3644119)
https://www.kaggle.com/zhangjuefei/birds-bones-and-living-habits
About
According to their living environments and living habits, birds are classified into different ecological groups. There are 8 ecological groups of birds:
- Swimming Birds
- Wading Birds
- Terrestrial Birds
- Raptors
- Scansorial Birds
- Singing Birds
- Cursorial Birds (not included in dataset)
- Marine Birds (not included in dataset)
First 6 groups are main and are covered by this dataset.
There are 420 birds contained in this dataset. Each bird is represented by 10 measurements (features):
- Length and Diameter of Humerus
- Length and Diameter of Ulna
- Length and Diameter of Femur
- Length and Diameter of Tibiotarsus
- Length and Diameter of Tarsometatarsus
# Load your packages
library(tidyr)
library(dplyr)
library(knitr)
library(GGally)
library(ggplot2)
library(ggridges)Data
# Load your data and prepare for visualisation
birdDS <- read.csv("bird.csv")Exploring the Data
# Load your data and prepare for visualisation
head(birdDS)colnames(birdDS)## [1] "id" "huml" "humw" "ulnal" "ulnaw" "feml" "femw" "tibl"
## [9] "tibw" "tarl" "tarw" "type"colNamesFull <- c("ID", "HumerusLength", "HumerusDiameter", "UlnaLength", "UlnaDiameter",
"FemurLength", "FemurDiameter", "TibiotarusLength", "TibiotarusDiameter",
"TarsometatarusLength", "TarsometatarusDiameter", "EcologicalBirdType")# Load your data and prepare for visualisation
summary(birdDS)## id huml humw ulnal
## Min. : 0.0 Min. : 9.85 Min. : 1.140 Min. : 14.09
## 1st Qu.:104.8 1st Qu.: 25.17 1st Qu.: 2.190 1st Qu.: 28.05
## Median :209.5 Median : 44.18 Median : 3.500 Median : 43.71
## Mean :209.5 Mean : 64.65 Mean : 4.371 Mean : 69.12
## 3rd Qu.:314.2 3rd Qu.: 90.31 3rd Qu.: 5.810 3rd Qu.: 97.52
## Max. :419.0 Max. :420.00 Max. :17.840 Max. :422.00
## NA's :1 NA's :1 NA's :3
## ulnaw feml femw tibl
## Min. : 1.000 Min. : 11.83 Min. : 0.930 Min. : 5.50
## 1st Qu.: 1.870 1st Qu.: 21.30 1st Qu.: 1.715 1st Qu.: 36.42
## Median : 2.945 Median : 31.13 Median : 2.520 Median : 52.12
## Mean : 3.597 Mean : 36.87 Mean : 3.221 Mean : 64.66
## 3rd Qu.: 4.770 3rd Qu.: 47.12 3rd Qu.: 4.135 3rd Qu.: 82.87
## Max. :12.000 Max. :117.07 Max. :11.640 Max. :240.00
## NA's :2 NA's :2 NA's :1 NA's :2
## tibw tarl tarw type
## Min. : 0.870 Min. : 7.77 Min. : 0.660 P : 38
## 1st Qu.: 1.565 1st Qu.: 23.04 1st Qu.: 1.425 R : 50
## Median : 2.490 Median : 31.74 Median : 2.230 SO:128
## Mean : 3.182 Mean : 39.23 Mean : 2.930 SW:116
## 3rd Qu.: 4.255 3rd Qu.: 50.25 3rd Qu.: 3.500 T : 23
## Max. :11.030 Max. :175.00 Max. :14.090 W : 65
## NA's :1 NA's :1 NA's :1# Load your data and prepare for visualisation
str(birdDS)## 'data.frame': 420 obs. of 12 variables:
## $ id : int 0 1 2 3 4 5 6 7 8 9 ...
## $ huml : num 80.8 88.9 80 77.7 62.8 ...
## $ humw : num 6.68 6.63 6.37 5.7 4.84 ...
## $ ulnal: num 72 80.5 69.3 65.8 52.1 ...
## $ ulnaw: num 4.88 5.59 5.28 4.77 3.73 3.47 4.5 4.55 6.13 7.05 ...
## $ feml : num 41.8 47 43.1 40 34 ...
## $ femw : num 3.7 4.3 3.9 3.52 2.72 4.41 3.41 3.78 5.45 7.44 ...
## $ tibl : num 5.5 80.2 75.3 69.2 56.3 ...
## $ tibw : num 4.03 4.51 4.04 3.4 2.96 2.73 3.56 3.81 5.58 7.31 ...
## $ tarl : num 38.7 41.5 38.3 35.8 31.9 ...
## $ tarw : num 3.84 4.01 3.34 3.41 3.13 2.83 3.64 3.81 4.37 6.34 ...
## $ type : Factor w/ 6 levels "P","R","SO","SW",..: 4 4 4 4 4 4 4 4 4 4 ...# Load your data and prepare for visualisation
head(birdDS)# Check the species ID
unique(birdDS$type)## [1] SW W T R P SO
## Levels: P R SO SW T WDetermine the mean values
# SW - Swimming Birds
SWbirds <- subset(birdDS, birdDS$type == "SW") # SW - Swimming Birds
Wbirds <- subset(birdDS, birdDS$type == "W") # W - Wading Birds
Tbirds <- subset(birdDS, birdDS$type == "T") # T - Terrestrial Birds
Rbirds <- subset(birdDS, birdDS$type == "R") # R - Raptors
Pbirds <- subset(birdDS, birdDS$type == "P") # P - Scansorial Birds
SObirds <- subset(birdDS, birdDS$type == "SO") # SO - Singing Birds# Filling NA - Rbirds
Wbirds$feml[is.na(Wbirds$feml)] <- mean(Wbirds$feml, na.rm = TRUE)
Wbirds$femw[is.na(Wbirds$femw)] <- mean(Wbirds$femw, na.rm = TRUE)# Filling NA - Rbirds
Rbirds$ulnal[is.na(Rbirds$ulnal)] <- mean(Rbirds$ulnal, na.rm = TRUE)
Rbirds$ulnaw[is.na(Rbirds$ulnaw)] <- mean(Rbirds$ulnaw, na.rm = TRUE)
Rbirds$tarl[is.na(Rbirds$tarl)] <- mean(Rbirds$tarl, na.rm = TRUE)
Rbirds$tarw[is.na(Rbirds$tarw)] <- mean(Rbirds$tarw, na.rm = TRUE)# Filling NA - SObirds
SObirds$huml[is.na(SObirds$huml)] <- mean(SObirds$huml, na.rm = TRUE)
SObirds$humw[is.na(SObirds$humw)] <- mean(SObirds$humw, na.rm = TRUE)
SObirds$ulnal[is.na(SObirds$ulnal)] <- mean(SObirds$ulnal, na.rm = TRUE)
SObirds$ulnaw[is.na(SObirds$ulnaw)] <- mean(SObirds$ulnaw, na.rm = TRUE)
SObirds$feml[is.na(SObirds$feml)] <- mean(SObirds$feml, na.rm = TRUE)
SObirds$tibl[is.na(SObirds$tibl)] <- mean(SObirds$tibl, na.rm = TRUE)
SObirds$tibw[is.na(SObirds$tibw)] <- mean(SObirds$tibw, na.rm = TRUE)birdDS_clean <- rbind(SWbirds, Wbirds, Tbirds, Rbirds, Pbirds, SObirds)summary(birdDS_clean)## id huml humw ulnal
## Min. : 0.0 Min. : 9.85 Min. : 1.140 Min. : 14.09
## 1st Qu.:104.8 1st Qu.: 25.04 1st Qu.: 2.188 1st Qu.: 28.00
## Median :209.5 Median : 44.08 Median : 3.495 Median : 43.51
## Mean :209.5 Mean : 64.55 Mean : 4.365 Mean : 68.99
## 3rd Qu.:314.2 3rd Qu.: 90.22 3rd Qu.: 5.805 3rd Qu.: 97.56
## Max. :419.0 Max. :420.00 Max. :17.840 Max. :422.00
## ulnaw feml femw tibl
## Min. : 1.000 Min. : 11.83 Min. : 0.930 Min. : 5.50
## 1st Qu.: 1.867 1st Qu.: 21.33 1st Qu.: 1.718 1st Qu.: 36.34
## Median : 2.945 Median : 31.13 Median : 2.525 Median : 51.94
## Mean : 3.596 Mean : 36.84 Mean : 3.221 Mean : 64.53
## 3rd Qu.: 4.772 3rd Qu.: 47.10 3rd Qu.: 4.122 3rd Qu.: 82.87
## Max. :12.000 Max. :117.07 Max. :11.640 Max. :240.00
## tibw tarl tarw type
## Min. : 0.870 Min. : 7.77 Min. : 0.660 P : 38
## 1st Qu.: 1.560 1st Qu.: 23.04 1st Qu.: 1.427 R : 50
## Median : 2.490 Median : 31.81 Median : 2.235 SO:128
## Mean : 3.178 Mean : 39.28 Mean : 2.935 SW:116
## 3rd Qu.: 4.253 3rd Qu.: 50.46 3rd Qu.: 3.522 T : 23
## Max. :11.030 Max. :175.00 Max. :14.090 W : 65Visualisation
colnames(birdDS_clean)## [1] "id" "huml" "humw" "ulnal" "ulnaw" "feml" "femw" "tibl"
## [9] "tibw" "tarl" "tarw" "type"allMeasurements <- c("huml", "humw", "ulnal", "ulnaw", "feml", "femw", "tibl", "tibw", "tarl", "tarw")
lengthMeasurements <- c("huml", "ulnal", "feml", "tibl", "tarl")
widthMeasurements <- c("humw", "ulnaw", "femw", "tibw", "tarw")test_reorder <- birdDS_clean[,c(12,2,3,4,5,6,7,8,9,10,11,1)]test <- birdDS_clean %>% gather("huml", "humw", "ulnal", "ulnaw", "feml", "femw", "tibl", "tibw", "tarl", "tarw",
key = "id", value = "measurement")SWbirdsLong <- SWbirds %>% gather("huml", "humw", "ulnal", "ulnaw", "feml", "femw", "tibl", "tibw", "tarl", "tarw",
key = "type", value = "measurement")
SWbirdsLong$id <- 'SW'
WbirdsLong <- Wbirds %>% gather("huml", "humw", "ulnal", "ulnaw", "feml", "femw", "tibl", "tibw", "tarl", "tarw",
key = "type", value = "measurement")
WbirdsLong$id <- 'W'
TbirdsLong <- Tbirds %>% gather("huml", "humw", "ulnal", "ulnaw", "feml", "femw", "tibl", "tibw", "tarl", "tarw",
key = "type", value = "measurement")
TbirdsLong$id <- 'T'
RbirdsLong <- Rbirds %>% gather("huml", "humw", "ulnal", "ulnaw", "feml", "femw", "tibl", "tibw", "tarl", "tarw",
key = "type", value = "measurement")
RbirdsLong$id <- 'R'
PbirdsLong <- Pbirds %>% gather("huml", "humw", "ulnal", "ulnaw", "feml", "femw", "tibl", "tibw", "tarl", "tarw",
key = "type", value = "measurement")
PbirdsLong$id <- 'P'
SObirdsLong <- SObirds %>% gather("huml", "humw", "ulnal", "ulnaw", "feml", "femw", "tibl", "tibw", "tarl", "tarw",
key = "type", value = "measurement")
SObirdsLong$id <- 'SO'birdDS_Long <- rbind(SWbirdsLong, WbirdsLong, TbirdsLong, RbirdsLong, PbirdsLong, SObirdsLong)summary(birdDS_Long)## id type measurement
## Length:4200 Length:4200 Min. : 0.66
## Class :character Class :character 1st Qu.: 2.67
## Mode :character Mode :character Median : 12.47
## Mean : 29.15
## 3rd Qu.: 38.50
## Max. :422.00colnames(birdDS_Long) <- c("birdType", "bone", "measurement")birdDS_Long$measurementType <- sapply(strsplit(as.character(birdDS_Long$bone), ""), tail, 1)
birdDS_Long$boneAlone = substr(birdDS_Long$bone,1,nchar(birdDS_Long$bone)-1)birdDS_Long$boneAlone <- factor(birdDS_Long$boneAlone)
levels(birdDS_Long$boneAlone) <- c("Femur", "Humerus", "Tarsometatarus", "Tibiotarsus", "Ulna")
levels(birdDS_Long$boneAlone)## [1] "Femur" "Humerus" "Tarsometatarus" "Tibiotarsus"
## [5] "Ulna"png('birds.png', units="in", width=32, height=18, res=300)
birdsPlot <- ggplot(data = birdDS_Long,
aes(x=birdDS_Long$measurement,
y = birdType,
fill = measurementType))
birdsPlot <- birdsPlot + geom_density_ridges2(scale = 0.9, alpha=0.7, colour = 'white', size = 0.5) +
labs(y = NULL, x = "Measurement (mm)",
title = "Bird Bone Measurements",
subtitle = "Alistair Grevis-James s3644119",
caption="Source: kaggle.com/zhangjuefei/birds-bones-and-living-habits.") +
theme(plot.title=element_text(size=62,
face="bold",
family="American Typewriter",
color="grey30"),
plot.subtitle=element_text(size=30,
family="American Typewriter",
color="grey30"),
plot.caption = element_text(size=24,
color="grey30"),
axis.title.x=element_text(vjust=0,
face = "bold",
family="American Typewriter",
size=30),
axis.text.y = element_text(vjust = 0,
family="American Typewriter",
size = 30),
axis.text.x = element_text(size = 20,
family="American Typewriter"),
strip.text.x = element_text(size = 30,
family="American Typewriter"),
legend.text = element_text(size=24)) +
guides(fill = guide_legend(keywidth = 2.5, keyheight = 2.5)) +
scale_x_continuous(expand = c(0.01, 0), breaks = c(0, 25, 50, 75, 100, 125)) +
coord_cartesian(xlim = c(0, 150)) +
scale_y_discrete(expand = c(0.01, 0), labels=c("Scansorial Birds", "Raptors",
"Song Birds", "Swimming Birds",
"Terrestrial Birds", "Wading Birds")) +
facet_grid(.~boneAlone) +
theme(legend.title = element_blank()) +
theme(legend.position = c(0.95, 0.09)) +
scale_fill_manual(values = c("#ff7f00", "#377eb8"), labels=c("Length", "Width"))
birdsPlot
dev.off()## quartz_off_screen
## 2birdsPlot
The figure shows a series of faceted density plots of bird bone measurements. The measurements are of either length in orange or diameter (width) in blue – all measurements are in mm. The plots are faceted by bone type, and each row represents a type of bird. The visualisation is highly effective at showing the relation between length and diameter of a particular bone, within a particular bird type. The visualisation also allows for distribution comparisons between bird groups and for distribution comparisons between bone types. Starting in the top left-hand corner, we see for example that femur measurements of wading birds are very widely distributed. If we look to terrestrial birds, we see the femur length and width are extremely highly correlated. This correlation can then be compared with the other bones measured.