1 Project Description
Tidy Tuesday has a weekly data project aimed at the R ecosystem. An emphasis is placed on understanding how to summarize and arrange data to make meaningful charts with ggplot2, tidyr, dplyr, and other tools in the tidyverse ecosystem.
2 Dataset
Data came from Bird Studies Canada. The data includes meteorite information such as year, bird species, and number of birds.
3 Setup
3.1 Load Libraries
#function to check if packages are installed, if not then install them, and load all packages
libraries <- function(packages){
for(package in packages){
#checks if package is installed
if(!require(package, character.only = TRUE)){
#If package does not exist, then it will install
install.packages(package, dependencies = TRUE)
#Loads package
library(package, character.only = TRUE)
}
}
}
packages <- c("data.table","tidyverse","visdat","zoo","extrafont","stringr")
libraries(packages)
theme_set(theme_classic())3.2 Import Data
df <- fread("https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2019/2019-06-18/bird_counts.csv", stringsAsFactors = TRUE)4 Exploratory Data Analysis
- There are 18,706 observations and 6 variables.
- how_many_counted_by_hour is how_many_counted divided by total_hours.
- There are many NA values with total_hours which results in NAs in how_many_counted_by_hour.
4.1 View Data
glimpse(df)## Observations: 18,706
## Variables: 6
## $ year <int> 1921, 1921, 1921, 1921, 1921, 1921, 1...
## $ species <fct> American Bittern, American Black Duck...
## $ species_latin <fct> Botaurus lentiginosus, Anas rubripes,...
## $ how_many_counted <int> 0, 0, 0, 0, 0, 0, 0, 1, 0, 12, 0, 0, ...
## $ total_hours <dbl> 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8, 8...
## $ how_many_counted_by_hour <dbl> 0.000, 0.000, 0.000, 0.000, 0.000, 0....head(df)## year species species_latin how_many_counted
## 1: 1921 American Bittern Botaurus lentiginosus 0
## 2: 1921 American Black Duck Anas rubripes 0
## 3: 1921 American Coot Fulica americana 0
## 4: 1921 American Crow Corvus brachyrhynchos 0
## 5: 1921 American Goldfinch Spinus tristis 0
## 6: 1921 American Kestrel Falco sparverius 0
## total_hours how_many_counted_by_hour
## 1: 8 0
## 2: 8 0
## 3: 8 0
## 4: 8 0
## 5: 8 0
## 6: 8 0summary(df)## year species species_latin
## Min. :1921 American Bittern : 94 Acanthis flammea : 94
## 1st Qu.:1947 American Black Duck: 94 Acanthis hornemanni: 94
## Median :1970 American Coot : 94 Accipiter cooperii : 94
## Mean :1970 American Crow : 94 Accipiter gentilis : 94
## 3rd Qu.:1994 American Goldfinch : 94 Accipiter striatus : 94
## Max. :2017 American Kestrel : 94 Actitis macularius : 94
## (Other) :18142 (Other) :18142
## how_many_counted total_hours how_many_counted_by_hour
## Min. : 0.0 Min. : 8.0 Min. : 0.000
## 1st Qu.: 0.0 1st Qu.:149.5 1st Qu.: 0.000
## Median : 0.0 Median :171.0 Median : 0.000
## Mean : 193.5 Mean :170.8 Mean : 1.336
## 3rd Qu.: 5.0 3rd Qu.:203.8 3rd Qu.: 0.051
## Max. :73000.0 Max. :251.0 Max. :439.024
## NA's :3781 NA's :3781sapply(df, function(x) n_distinct(x)) %>% sort()## total_hours year species
## 62 94 199
## species_latin how_many_counted how_many_counted_by_hour
## 199 1163 35564.2 Missing Values
View missing values in more detail.
#Visualize missing values
vis_miss(df, sort_miss=TRUE)
#see count of missing values
na_values <- function(df){
na <- colSums(is.na(df)) %>% sort(decreasing=TRUE)
na[na>0]
}
na_values(df)## total_hours how_many_counted_by_hour
## 3781 37814.3 Visualize Trends
df %>% ggplot(aes(year,(how_many_counted/total_hours)))+
geom_bar(stat="identity")+
labs(title="Christmas Bird Counts Per Hour Over Time", x="Year", y="Count per Hour")
df %>% ggplot(aes(total_hours, how_many_counted))+
geom_jitter(alpha=0.3, size=1)+
labs(title="Total Christmas Bird Counts",x="Hours", y="Count")
5 Data Wrangling
Remove/replace missing values and drop unnecessary columns
#remove NA hours if how_many_counted is 0
df <- df[!(df$how_many_counted==0 & is.na(df$total_hours)),]
#view missing data again
na_values(df)## total_hours how_many_counted_by_hour
## 753 753# use interpolated values to replace NA values, grouped by species
df <- df %>% group_by(species) %>% mutate(total_hours = na.approx(total_hours)) %>% ungroup()
#view missing data again
na_values(df)## how_many_counted_by_hour
## 753#drop how_many_counted_by_hour. this can be recalculated later
df <- df %>% select(1:5)6 Final Visualization
# how many owls were seen
df_owl <- df %>% filter(str_detect(species, "Owl"))
df_owl <- df_owl %>% group_by(species) %>% summarize(counted = sum(how_many_counted))
df_owl <- df_owl %>% mutate(species = str_sub(species, end=-4))
df_owl$species <- fct_reorder(df_owl$species, df_owl$counted)
df_owl %>% ggplot(aes(species,counted))+
geom_col(fill="steelblue")+
geom_text(aes(label=counted), hjust=-0.1, vjust=0.5, family="Bodoni MT", fontface="bold")+
coord_flip()+
labs(title="Christmas Owl Spottings", subtitle="From 1921 - 2017", y="Number Spotted")+
theme(legend.position="none",
text=element_text(family="Bodoni MT", size=14),
axis.title.y=element_blank(),
plot.title=element_text(face="bold", size=24, hjust=0.175),
plot.subtitle=element_text(hjust=0.325),
plot.background=element_rect(fill="#fff7e6"),
panel.background=element_rect(fill="#fff7e6"))
ggsave("christmas_owls.png", limitsize=FALSE)## Saving 7 x 5 in image