Spinning Cycles in Box #4 To Take the Pies out of Pi Day
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.
I caught this tweet today:
INSTEAD OF WORRYING ABOUT HOW OTHER DATA SCIENTISTS USE
— Metrics Hulk (@MetricsHulk) March 14, 2016PERHAPS @WSJGraphics COULD SPEND THEIR
The WSJ folks usually do a great job, but this was either rushed or not completely thought through. There’s no way you’re going to be able to do any real comparisons between the segments across pies and direct pie % labels kinda mean they should have just made a table if they were going to phone it in.
Despite the fact that today is Pi[e] Day, these pies need to go.
If the intent was to primarily allow comparison of hours in-task, leaving some ability to compare the same time category across tasks, then bars are probably the way to go (you could do a parallel coordinates plot, but those looks like tangled guitar strings to me, so I’ll stick with bars). Here’s one possible alternative using R & ggplot2. Since I provide the data, please link to your own creations as I’d love to see how others would represent the data.
NOTE: I left direct bar labels off deliberately. My view is that (a) this is designed to be a relative comparison vs precise comparison & (b) it’s survey data and if we’re going to add #’s I’d feel compelled to communicate margin of error, etc. I don’t think that’s necessary.
library(ggplot2)
library(grid)
library(scales)
library(hrbrmisc) # devtools::install_github("hrbrmstr/hrbrmisc")
library(tidyr)
Â
dat <- read.table(text=
"Task|less_than_one_hour_per_week|one_to_four_hours_per_week|one_to_three_hours_a_day|four_or_more_hours_a_day
Basic exploratory data analysis|11|32|46|12
Data cleaning|19|42|31|7
Machine learning, statistics|34|29|27|10
Creating visualizations|23|41|29|7
Presenting analysis|27|47|20|6
Extract, transform, load|43|32|20|5", sep="|", header=TRUE, stringsAsFactors=FALSE)
Â
amount_trans <- c("one_to_four_hours_per_week"="1-4 hrs/nwk",
"less_than_one_hour_per_week"="<1 hr/nwk",
"one_to_three_hours_a_day"="1-3 hrs/nday",
"four_or_more_hours_a_day"="4+ hrs/nday")
Â
dat <- gather(dat, amount, value, -Task)
dat$value <- dat$value / 100
dat$amount <- factor(amount_trans[dat$amount], levels=amount_trans)
Â
title_trans <- c("Basic exploratory data analysis"="Basic exploratoryndata analysis",
"Data cleaning"="Datancleaning",
"Machine learning, statistics"="Machine learning,nstatistics",
"Creating visualizations"="Creatingnvisualizations",
"Presenting analysis"="Presentingnanalysis",
"Extract, transform, load"="Extract,ntransform, load")
Â
dat$Task <-factor(title_trans[dat$Task], levels=title_trans)
Â
gg <- ggplot(dat, aes(x=amount, y=value, fill=amount))
gg <- gg + geom_bar(stat="identity", width=0.75, color="#2b2b2b", size=0.05)
gg <- gg + scale_y_continuous(expand=c(0,0), labels=percent, limits=c(0, 0.5))
gg <- gg + scale_x_discrete(expand=c(0,1))
gg <- gg + scale_fill_manual(name="", values=c("#a6cdd9", "#d2e4ee", "#b7b079", "#efc750"))
gg <- gg + facet_wrap(~Task)
gg <- gg + labs(x=NULL, y=NULL, title="Where Does the Time Go?")
gg <- gg + theme_hrbrmstr(grid="Y", axis="x", plot_title_margin=9)
gg <- gg + theme(panel.background=element_rect(fill="#efefef", color=NA))
gg <- gg + theme(strip.background=element_rect(fill="#858585", color=NA))
gg <- gg + theme(strip.text=element_text(family="OpenSans-CondensedBold", size=12, color="white", hjust=0.5))
gg <- gg + theme(panel.margin.x=unit(1, "cm"))
gg <- gg + theme(panel.margin.y=unit(0.5, "cm"))
gg <- gg + theme(legend.position="none")
gg <- gg + theme(panel.grid.major.y=element_line(color="#b2b2b2"))
gg <- gg + theme(axis.text.x=element_text(margin=margin(t=-10)))
gg <- gg + theme(axis.text.y=element_text(margin=margin(r=-10)))
Â
ggplot_with_subtitle(gg,
"The amount of time spent on various tasks by surveyed non-managers in data-science positions.",
fontfamily="OpenSans-CondensedLight", fontsize=12, bottom_margin=16) |
R-bloggers.com offers daily e-mail updates about R news and tutorials about learning R and many other topics. Click here if you're looking to post or find an R/data-science job.
Want to share your content on R-bloggers? click here if you have a blog, or here if you don't.