dataviz

Research methods classes in graduate school generally teach important skills such as probability and statistical theory, regression, analysis of variance (ANOVA), maximum likelihood estimation (MLE), etc. While these are important methods for analyzing data and assessing research questions, sometimes drawing a picture (aka visualization) can be more precise than conventional statistical computations.

This page is a summary of A Layered Grammar of Graphics by Hadley Wickham. I strongly encourage you to read the original article in conjunction with this summary. library(tidyverse) library(knitr) library(palmerpenguins) Google defines a grammar as “the whole system and structure of a language or of languages in general, usually taken as consisting of syntax and morphology (including inflections) and sometimes also phonology and semantics”.

library(tidyverse) library(knitr) library(here) Figure 1: Charles Minard's 1869 chart showing the number of men in Napoleon’s 1812Russian campaign army, their movements, as well as the temperature they encounteredon the return path.

library(tidyverse) Run the code below in your console to download this exercise as a set of R scripts. usethis::use_course("cis-ds/grammar-of-graphics") Given your preparation for today’s class, now let’s practice generating layered graphics in R using data from Gapminder World, which compiles country-level data on quality-of-life measures.

Geospatial visualizations are one of the earliest forms of information visualizations. They were used historically for navigation and were essential tools before the modern technological era of humanity. Data maps were first popularized in the seventeenth century and have grown in complexity and detail since then.

library(tidyverse) library(ggmap) library(RColorBrewer) library(patchwork) library(here) options(digits = 3) set.seed(1234) theme_set(theme_minimal()) ggmap is a package for R that retrieves raster map tiles from online mapping services like Stamen Maps and plots them using the ggplot2 framework.

library(tidyverse) library(ggmap) library(here) options(digits = 3) set.seed(123) theme_set(theme_minimal()) Run the code below in your console to download this exercise as a set of R scripts. usethis::use_course("cis-ds/visualize-spatial-i") New York City 311 data New York City has an excellent data portal publishing a large volume of public records.

library(tidyverse) library(sf) library(here) options(digits = 3) set.seed(1234) theme_set(theme_minimal()) Rather than storing spatial data as raster image files which are not easily modifiable, we can instead store spatial data as vector files.

library(tidyverse) library(sf) library(here) options(digits = 3) set.seed(1234) theme_set(theme_minimal()) Unlike raster image maps, vector maps require you to obtain spatial data files which contain detailed information necessary to draw all the components of a map (e.

library(tidyverse) library(sf) library(tidycensus) library(colorspace) library(scales) # useful on MacOS to speed up rendering of geom_sf() objects if (!identical(getOption("bitmapType"), "cairo") && isTRUE(capabilities()[["cairo"]])) { options(bitmapType = "cairo") } options(digits = 3) set.seed(123) theme_set(theme_minimal()) Run the code below in your console to download this exercise as a set of R scripts.

library(tidyverse) library(sf) library(tidycensus) library(RColorBrewer) library(patchwork) # useful on MacOS to speed up rendering of geom_sf() objects if (!identical(getOption("bitmapType"), "cairo") && isTRUE(capabilities()[["cairo"]])) { options(bitmapType = "cairo") } options(digits = 3) set.seed(1234) theme_set(theme_minimal()) Selection of your color palette is perhaps the most important decision to make when drawing a choropleth.