library(dplyr)
CO2 %>%
mutate(uptake = case_when(
uptake > 12 ~ "yay!",
uptake > 10 ~ "OK.",
TRUE ~ "oh no..."
))UQRUG 10
meeting
visualisation
dplyr
ggplot2
visualisation
Main: TidyTuesday.
Other topics: dplyr:case_when() to recode, working directory, ggplot2 labels and limits, epiDisplay, lme4, Seurat, reshape2, lattice.
Other topics: dplyr:case_when() to recode, working directory, ggplot2 labels and limits, epiDisplay, lme4, Seurat, reshape2, lattice.
2020-10-19: UQRUG 10
Attendees and problems
Please add your name and the problem you’d like some help with!
- Stéphane (Library): here to help!
- Chantelle: Biological Sciences. I can’t stay to long as I have to go track carpet pythons at 4pm ;)
- Einat (Social Sciences, R studio beginner): Apologies in advance would have to leave early. have some issues wiTh minor and basic procedures, like …CONDITIONS!! and also with reading labels from csv/excel file.
- Patrick (Agricultural Science) Honours student
- Gabriel: Hydrologist, spatial data analyst. PhD candidate at SMI. R beginner and Python enthusiast !
- Natalie (Mathematics): Undergraduate Mathematics and Statistics student
- Phoebe: learn more about data viz!
- … and few more!
Today’s challenge
Try playing with the latest TidyTuesday dataset: https://github.com/rfordatascience/tidytuesday/blob/master/data/2020/2020-10-13/readme.md
Share your findings and cool visualisations with the group!
Recoding data
This is a working example of recoding using dplyr’s case_when():
The TRUE line is for “everything else”.
Finding files
The “working directory” is the default location:
getwd() # know where you are
setwd("path/to/correct/location") # change it
# functions will look there by default
read.csv("filename.csv")
# but you can use longer, absolute filepaths
read.csv("C:/Users/myname/filename.csv")If you don’t want to deal with that: use R Projects! Creating an R Project will set the working directory for you.
Dealing with overflowing labels
library(ggplot2)
ggplot(diamonds, aes(x = cut)) +
geom_bar()
# horizontal bar chart
ggplot(diamonds, aes(y = cut)) +
geom_bar()
# abbreviate the labels
abbreviate(diamonds$cut)
# using it in the visualisation is nice and concise
ggplot(diamonds, aes(x = cut)) +
geom_bar() +
scale_x_discrete(labels = abbreviate)On the other hand, if labels are overflowing inside the canvas, we can expand axis limits:
# expand limits
ggplot(diamonds, aes(x = cut)) +
geom_bar() +
ylim(c(0, 25000))However, ggplot2 should expand the plot area to include all the geometries automatically, including a geom_label() or geom_text().