class: inverse background-image: url("img/PicAnim.gif") # Leaders and Grammar for Graphics --- # A Grammar of Graphics -- - data -- - a mapping from data to aesthetics -- - a geometry for representing the aesthetic --- class: inverse # My Primitive: A canvas -- ### Data as coordinates in 1 or 2-d with aesthetics and geometries -- Graphics are generically flat; they are (at most) **two dimensional**. That is caused by the screen. -- **Our visuals must respect these limitations but we can add colors, shapes, and sizes,** -- or even emojis. 🔥  --- class:inverse, center, middle, bounceIn, bounceOut # The canvas: Every graphic fits here <img src="index_files/figure-html/unnamed-chunk-1-1.svg" style="display: block; margin: auto;" /> --- class: inverse, tada # One Important Caveat There is one notable and widely seen exception to this charaterization. Size is total tenure.<sup>1<sup> .footnote[ [1] The install script for the packages used is in the talk directory as `source(url("http://rww.science/talk/ggsc/PackagesUsed.R"))`. The interested reader can issue `source(url("https://rww.science/talk/ggsc/XSCSetup.R"))`. ] --- class: inverse,heartBeat ## The Illustration What determines x and y? **Nothing.** Coordinates are arbitrary.  --- class: inverse, center, middle # A Simple Example to Cement our Thinking -- [Where am I?](https://where-am-i.net) --- # I am in Oregon ## Let's Start There Get the map of Oregon and then use `ggmap` to map it. ```r OR_map <- get_map(getbb("Oregon")) # ggmap(OR_map) ``` --- class: inverse, wobble ## The Result <!-- --> --- class: inverse # Me as Data For now, just my GPS coordinates. ```r Me <- data.frame(lat=44.924943, lon=-123.043461) ```  --- # Mapping Me... - Put my data on the earlier map. - My data is a set of coordinates that are already defined on the map. - Let's locate me on the map. ```r Me <- data.frame(lat=44.924943, lon=-123.043461) ggmap(OR_map) + geom_emoji(data=Me, aes(y=lat, x=lon), emoji="1f4aa", size=0.125) + theme_nothing() ``` --- class: inverse # Me on the Map <!-- --> --- class: inverse, heartBeat # Data Visualisation on the Canvas ## Showcasing the Language of Visuals with Data on National Leaders --- # The Data # Archigos: A Database of National Leaders Chiozza, Gleditsch, and Goemans have collected and maintained an encyclopedia of world leaders since the 1870s. -- - Leader spells are units -- - Nested in countries -- - Entry and exit including types -- - Real date formats 😠```r Archigos <- read_dta(url("http://www.rochester.edu/college/faculty/hgoemans/Archigos_4.1_stata14.dta")) Archigos <- Archigos %>% mutate(leader = utf8_encode(leader)) kable(Archigos) %>% kable_styling() %>% scroll_box(width = "800px", height = "400px") ``` .footnote[ [1] Archigos [on the web](http://www.ksgleditsch.com/archigos.html). ] --- ## A Look <div style="border: 1px solid #ddd; padding: 0px; overflow-y: scroll; height:400px; overflow-x: scroll; width:800px; "><table class="table" style="margin-left: auto; margin-right: auto;"> <thead> <tr> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> obsid </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> leadid </th> <th style="text-align:right;position: sticky; top:0; background-color: #FFFFFF;"> ccode </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> idacr </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> leader </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> startdate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> eindate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> enddate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> eoutdate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> entry </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> exit </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> exitcode </th> <th style="text-align:right;position: sticky; top:0; background-color: #FFFFFF;"> prevtimesinoffice </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> posttenurefate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> gender </th> <th style="text-align:right;position: sticky; top:0; background-color: #FFFFFF;"> yrborn </th> <th style="text-align:right;position: sticky; top:0; background-color: #FFFFFF;"> yrdied </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> borndate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> ebirthdate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> deathdate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> edeathdate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> dbpediauri </th> <th style="text-align:right;position: sticky; top:0; background-color: #FFFFFF;"> numentry </th> <th style="text-align:right;position: sticky; top:0; background-color: #FFFFFF;"> numexit </th> <th style="text-align:right;position: sticky; top:0; background-color: #FFFFFF;"> numexitcode </th> <th style="text-align:right;position: sticky; top:0; background-color: #FFFFFF;"> numposttenurefate </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> fties </th> <th style="text-align:left;position: sticky; top:0; background-color: #FFFFFF;"> ftcur </th> </tr> </thead> <tbody> <tr> <td style="text-align:left;"> USA-1869 </td> <td style="text-align:left;"> 81dcc176-1e42-11e4-b4cd-db5882bf8def </td> <td style="text-align:right;"> 2 </td> <td style="text-align:left;"> USA </td> <td style="text-align:left;"> Grant </td> <td style="text-align:left;"> 1869-03-04 </td> <td style="text-align:left;"> 1869-03-04 </td> <td style="text-align:left;"> 1877-03-04 </td> <td style="text-align:left;"> 1877-03-04 </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> OK </td> <td style="text-align:left;"> M </td> <td style="text-align:right;"> 1822 </td> <td style="text-align:right;"> 1885 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> 1885-07-23 </td> <td style="text-align:left;"> 1885-07-23 </td> <td style="text-align:left;"> https://urldefense.proofpoint.com/v2/url?u=http-3A__dbpedia.org_resource_Ulysses-5FS.-5FGrant&d=BQID-g&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=ZliGVSRwfirWoETOrKCh2RSnoygzVPWEk95Me9L-Kwo&m=EaKyFx9mpkhPdFxsxzvW_fiM3jbYwM3xYLjVbSFoDAg&s=x7yRI4mbwG8b0j8c_AwqAahtkPcCfIY2uH8tdRaReBw&e= </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 1 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> </tr> <tr> <td style="text-align:left;"> USA-1877 </td> <td style="text-align:left;"> 81dcc177-1e42-11e4-b4cd-db5882bf8def </td> <td style="text-align:right;"> 2 </td> <td style="text-align:left;"> USA </td> <td style="text-align:left;"> Hayes </td> <td style="text-align:left;"> 1877-03-04 </td> <td style="text-align:left;"> 1877-03-04 </td> <td style="text-align:left;"> 1881-03-04 </td> <td style="text-align:left;"> 1881-03-04 </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> OK </td> <td style="text-align:left;"> M </td> <td style="text-align:right;"> 1822 </td> <td style="text-align:right;"> 1893 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> 1893-01-17 </td> <td style="text-align:left;"> 1893-01-17 </td> <td style="text-align:left;"> https://urldefense.proofpoint.com/v2/url?u=http-3A__dbpedia.org_resource_Rutherford-5FB.-5FHayes&d=BQID-g&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=ZliGVSRwfirWoETOrKCh2RSnoygzVPWEk95Me9L-Kwo&m=EaKyFx9mpkhPdFxsxzvW_fiM3jbYwM3xYLjVbSFoDAg&s=8oshcxB52W-Pg-2a6AvRVJEYMU-S-RG55-369eJ-PyY&e= </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 1 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> </tr> <tr> <td style="text-align:left;"> USA-1881-1 </td> <td style="text-align:left;"> 81dcf24a-1e42-11e4-b4cd-db5882bf8def </td> <td style="text-align:right;"> 2 </td> <td style="text-align:left;"> USA </td> <td style="text-align:left;"> Garfield </td> <td style="text-align:left;"> 1881-03-04 </td> <td style="text-align:left;"> 1881-03-04 </td> <td style="text-align:left;"> 1881-09-19 </td> <td style="text-align:left;"> 1881-09-19 </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Irregular </td> <td style="text-align:left;"> Assassination by Unsupported Individual </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> Death </td> <td style="text-align:left;"> M </td> <td style="text-align:right;"> 1831 </td> <td style="text-align:right;"> 1881 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> 1881-09-19 </td> <td style="text-align:left;"> 1881-09-19 </td> <td style="text-align:left;"> https://urldefense.proofpoint.com/v2/url?u=http-3A__dbpedia.org_resource_James-5FA.-5FGarfield&d=BQID-g&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=ZliGVSRwfirWoETOrKCh2RSnoygzVPWEk95Me9L-Kwo&m=EaKyFx9mpkhPdFxsxzvW_fiM3jbYwM3xYLjVbSFoDAg&s=qTPSlRS-mPD_m8ZM3luuJ4W47DtmOPoXSnSZn-lYsqM&e= </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 3 </td> <td style="text-align:right;"> 11 </td> <td style="text-align:right;"> 3 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> </tr> <tr> <td style="text-align:left;"> USA-1881-2 </td> <td style="text-align:left;"> 81dcf24b-1e42-11e4-b4cd-db5882bf8def </td> <td style="text-align:right;"> 2 </td> <td style="text-align:left;"> USA </td> <td style="text-align:left;"> Arthur </td> <td style="text-align:left;"> 1881-09-19 </td> <td style="text-align:left;"> 1881-09-19 </td> <td style="text-align:left;"> 1885-03-04 </td> <td style="text-align:left;"> 1885-03-04 </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> OK </td> <td style="text-align:left;"> M </td> <td style="text-align:right;"> 1829 </td> <td style="text-align:right;"> 1886 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> 1886-11-18 </td> <td style="text-align:left;"> 1886-11-18 </td> <td style="text-align:left;"> https://urldefense.proofpoint.com/v2/url?u=http-3A__dbpedia.org_resource_Chester-5FA.-5FArthur&d=BQID-g&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=ZliGVSRwfirWoETOrKCh2RSnoygzVPWEk95Me9L-Kwo&m=EaKyFx9mpkhPdFxsxzvW_fiM3jbYwM3xYLjVbSFoDAg&s=XiglK4IJ3ZCC1kR4Hmb8eVB-V7A0tQprCtTKzOl42YM&e= </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 1 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> </tr> <tr> <td style="text-align:left;"> USA-1885 </td> <td style="text-align:left;"> 34fb1558-3bbd-11e5-afeb-eb6f07f9fec7 </td> <td style="text-align:right;"> 2 </td> <td style="text-align:left;"> USA </td> <td style="text-align:left;"> Cleveland </td> <td style="text-align:left;"> 1885-03-04 </td> <td style="text-align:left;"> 1885-03-04 </td> <td style="text-align:left;"> 1889-03-04 </td> <td style="text-align:left;"> 1889-03-04 </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> OK </td> <td style="text-align:left;"> M </td> <td style="text-align:right;"> 1837 </td> <td style="text-align:right;"> 1908 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> 1908-06-24 </td> <td style="text-align:left;"> 1908-06-24 </td> <td style="text-align:left;"> https://urldefense.proofpoint.com/v2/url?u=http-3A__dbpedia.org_resource_Grover-5FCleveland&d=BQID-g&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=ZliGVSRwfirWoETOrKCh2RSnoygzVPWEk95Me9L-Kwo&m=EaKyFx9mpkhPdFxsxzvW_fiM3jbYwM3xYLjVbSFoDAg&s=BH6RahZV58hS1fIGz1muHnQkbAirSz4Yy2wc6chejsQ&e= </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 1 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> </tr> <tr> <td style="text-align:left;"> USA-1889 </td> <td style="text-align:left;"> 81dcf24d-1e42-11e4-b4cd-db5882bf8def </td> <td style="text-align:right;"> 2 </td> <td style="text-align:left;"> USA </td> <td style="text-align:left;"> Harrison </td> <td style="text-align:left;"> 1889-03-04 </td> <td style="text-align:left;"> 1889-03-04 </td> <td style="text-align:left;"> 1893-03-04 </td> <td style="text-align:left;"> 1893-03-04 </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:left;"> Regular </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> OK </td> <td style="text-align:left;"> M </td> <td style="text-align:right;"> 1833 </td> <td style="text-align:right;"> 1901 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> 1901-03-13 </td> <td style="text-align:left;"> 1901-03-13 </td> <td style="text-align:left;"> https://urldefense.proofpoint.com/v2/url?u=http-3A__dbpedia.org_resource_Benjamin-5FHarrison&d=BQID-g&c=kbmfwr1Yojg42sGEpaQh5ofMHBeTl9EI2eaqQZhHbOU&r=ZliGVSRwfirWoETOrKCh2RSnoygzVPWEk95Me9L-Kwo&m=EaKyFx9mpkhPdFxsxzvW_fiM3jbYwM3xYLjVbSFoDAg&s=MA-TUrYNLJuZ9PTyDnsCkyBUtUyUkXWBHX7_p3f3pnA&e= </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 1 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:right;"> 0 </td> <td style="text-align:left;"> NA </td> <td style="text-align:left;"> NA </td> </tr> </tbody> </table></div> --- class: inverse, bottom, bounceIn background-image: url("img/WordCloudLeaders.jpeg") background-size: 500px background-position: 80% 8% # One discrete variable: ## Frequency [Barplot] or Probability [Mosaic] -- ```r Archigos %>% ggplot(., aes(x=leader)) + geom_bar(fill="#88398A") + labs(x="Leader", y="Frequency", title="The Frequency of Leaders", subtitle = "Leaders had as many as six tenures") ``` --- class: inverse ## too many values of x make bad graphics <!-- --> --- class: inverse, hinge ## Show More than Three Spells ```r Archigos %>% group_by(leader) %>% mutate(count = n()) %>% filter(count > 3) %>% ggplot(., aes(x=leader)) + geom_bar(fill="#88398A") + labs(x="Leader", y="Leadership Spells", title="The Frequency of Leaders", subtitle = "Leaders with more than three tenures") + coord_flip() ``` <!-- --> --- class: inverse # A Small Number of Categories Works Best -- How did leaders enter? ```r Archigos %>% ggplot(., aes(x=entry)) + geom_bar(fill="#88398A") + labs(title="How did the Leader Enter?", y="Count of Leaders", x="") + coord_flip() + theme_economist_white() ``` <!-- --> --- class: inverse # A Small Number of Categories Works Best -- How did leaders enter? [Probability] ```r Archigos %>% ggplot(.) + geom_mosaic(aes(x=product(entry), fill=entry)) + labs(title="How did the Leader Enter?", y="", x="", fill="Method of Entry") + coord_flip() ``` ``` ## Warning in grid.Call.graphics(C_rect, x$x, x$y, x$width, x$height, ## resolveHJust(x$just, : semi-transparency is not supported on this device: ## reported only once per page ``` <!-- --> --- class: inverse, center, middle, bounceOut # One quantitative variable: Leadership tenure ## How long do leaders stay in office? --- class: right background-image: url("img/TenureBack.svg") background-size: contain # Visualising Tenure: One numeric variable ## choose a scale -- ### connect points, create polygons, plot statistics --- class: inverse # One numeric variable: ### Visualising Tenure ```r Archigos <- Archigos %>% mutate(tenureY = year(enddate) - year(startdate)) plt1 <- Archigos %>% ggplot(., aes(tenureY)) + {{ geom_area(stat="bin", bins=30, fill="#88398A") }} + labs(x="Tenure (in years)", y="Frequency", title="geom_area") + theme_economist_white() plt2 <- Archigos %>% ggplot(., aes(tenureY)) + {{ geom_density(, fill="#88398A") }} + labs(x="Tenure (in years)", y="Frequency", title="geom_density") + theme_economist_white() plt3 <- Archigos %>% ggplot(., aes(tenureY)) + {{ geom_histogram(, fill="#88398A") }} + labs(x="Tenure (in years)", y="Frequency", title="geom_histogram") + theme_economist_white() plt4 <- Archigos %>% ggplot(., aes(tenureY)) + {{ geom_dotplot(, fill="#88398A") }} + labs(x="Tenure (in years)", y="Frequency", title="geom_dotplot") + theme_economist_white() grid.arrange(plt1,plt2,plt3,plt4) ``` --- class: inverse, rubberBand, center, middle <!-- --> --- class: inverse, zoomInDown ## Using the Percentiles Leads to a Boxplot ```r Archigos %>% ggplot(., aes(y=tenureY)) + geom_boxplot(fill="#88398A") + labs(x="Tenure (in years)", y="Frequency", title="geom_boxplot") + theme_economist_white() ``` <!-- --> --- class: inverse, flash, middle, center # Two Discrete Variables: ## A Table --- # How do leaders enter and exit? ```r Archigos %>% tabyl(exit,entry) ``` ``` ## exit Foreign Imposition Irregular Regular Unknown ## Foreign 8 22 42 0 ## Irregular 5 255 340 0 ## Natural Death 5 40 157 0 ## Regular 21 236 2036 0 ## Retired Due to Ill Health 0 5 63 0 ## Still in Office 0 12 156 0 ## Suicide 0 1 2 0 ## Unknown 1 0 0 2 ``` ```r Archigos %>% tabyl(exit,entry) %>% adorn_percentages("col") %>% adorn_pct_formatting() ``` ``` ## exit Foreign Imposition Irregular Regular Unknown ## Foreign 20.0% 3.9% 1.5% 0.0% ## Irregular 12.5% 44.7% 12.2% 0.0% ## Natural Death 12.5% 7.0% 5.6% 0.0% ## Regular 52.5% 41.3% 72.8% 0.0% ## Retired Due to Ill Health 0.0% 0.9% 2.3% 0.0% ## Still in Office 0.0% 2.1% 5.6% 0.0% ## Suicide 0.0% 0.2% 0.1% 0.0% ## Unknown 2.5% 0.0% 0.0% 100.0% ``` --- class:inverse # Visualizing Two Categorical Variables -- The Table is more informative because the *geometry* obscures information. ```r Archigos %>% ggplot(., aes(x=entry,y=exit)) + geom_count(color="purple") + labs(x="How did the Leader Enter?", y="How did the Leader Exit?", title="Leader Entry and Exit by Spell") + theme_economist() ``` <!-- --> --- class: inverse ## Geometries allow area (like maps) on a unit square It is usually suggested to avoid pie charts because squares are hard enough. ```r library(ggmosaic) ggplot(data=Archigos) + geom_mosaic(aes(x = product(exit,entry), fill=exit)) + labs(y="Method of Exit", x="Method of Entry", main="Entry and Exit by Leaders") + theme_minimal() + guides(fill=guide_legend(title="Method of Exit")) + theme(axis.text.y = element_text(size=8, hjust=0.25), axis.text.x = element_text(size=8, angle=45, vjust=1), legend.text = element_text(size=10)) ``` --- class: inverse, heartBeat <!-- --> --- class: inverse # Visualising Flows An alluvial shows how individuals pass through shared system points. For example, leaders enter, exit, and have some post tenure fate. --- class: inverse This needs work but I don't have a firm idea of how to collapse categories to improve readability. <!-- --> --- class: inverse, middle, center # One Discrete and one Quantitative Variable -- ## How does Tenure vary by Method of Entry? --- # The canvas - x is method of entry - y is tenure in years --- class:inverse # The geometry is everything ### Jitters Let me show the points [jittered]. <!-- --> --- # The single quantitative distributions don't work equally well. Five common types: - `geom_col` - `geom_dotplot` - `geom_boxplot` - `geom_violin` - `geom_ridges` --- class: inverse ## Common types: geom_col -- ### A grouped barplot **It sums the y values over x** ```r Archigos %>% ggplot(., aes(entry, tenureY)) + geom_col() + labs(y="Tenure (in years)", x="Method of Entry", title="geom_col") ``` <!-- --> --- class: inverse ## Common types: geom_dotplot -- ### A grouped dotplot How big are the bins, on what axis, and stacked how? They can run into each other ```r Archigos %>% ggplot(., aes(entry, tenureY)) + geom_dotplot(binwidth=1/20, binaxis = "y", stackdir = "center") + labs(y="Tenure (in years)", x="Entry Method", title="geom_dotplot") ``` <!-- --> --- class: inverse ## Common types: geom_boxplot -- ### A grouped boxplot ```r Archigos %>% ggplot(., aes(entry, tenureY)) + geom_boxplot() + labs(y="Tenure (in years)", x="Entry Method", title="geom_boxplot") ``` <!-- --> --- class: inverse ## Common types: geom_violin -- ### A grouped density projected in two dimensions ```r Archigos %>% ggplot(., aes(entry, tenureY)) + geom_violin() + labs(y="Tenure (in years)", x="Entry Method", title="geom_violin") ``` <!-- --> --- class: inverse, bounceIn ## Common types: geom_ridgeline -- ### A grouped density projected in two dimensions (flip x and y) ```r Archigos %>% ggplot(., aes(x=tenureY, y=entry, fill=factor(..quantile..))) + stat_density_ridges(geom = "density_ridges_gradient", calc_ecdf = TRUE, quantiles = 4, quantile_lines = TRUE, scale = 1) + scale_fill_viridis(discrete = TRUE, name = "Quartiles") + labs(x="Tenure (in years)", y="Entry Method", title="geom_ridges") ``` <!-- --> --- class: inverse, center, middle # Two Quantative Variables: -- ## Tenure and the Year Spells End --- background-image: url("img/Canvas.svg") background-size: contain # My canvas has been arbitrarily rendered. -- Now map the data to the canvas. - Replace x axis with Years Spells End - Replace y axis with Tenure --- class: background-image: url("img/Canvas-2.svg") background-size: contain # My canvas has coordinates. -- Geometries are:<sup>1<sup> - polygons [Oregon?] - **points** - connectors [lines, smooths] .footnote[ [1] The root cause is pixels. ] --- class: inverse, swing background-image: url("img/Canvas-2.svg") background-size: contain # Adding information with layers .pull-left[ Layers - alpha - color and/or fill - shape or label - size ] -- .pull-right[  ] --- class: inverse, flash # Tenure by Year Spells End It's messy. In many ways, that's perfect. - geom_point() makes it worse, they're stacked <!-- --> --- class: inverse background-image: url("img/ScatterLayer.svg") background-size: contain # Tenure by Year Spells End This is better but still messy. - Jitter points around a bit with geom_jitter() ```r Archigos %>% ggplot(aes(x=endYear, y=tenureY)) + * geom_jitter() + theme_dark() + labs(x="Year Leader's Spell Ended", y="Tenure of Leader (not cumulative)") ``` <!-- --> --- class: inverse background-image: url("img/JitterLayer.svg") background-size: contain # Layers -- .pull-left[ Color aesthetics: - **alpha** - alpha controls transparency - concentration can be seen ```r Archigos %>% ggplot(aes(x=endYear, y=tenureY)) + * geom_jitter(alpha=0.1) + theme_economist_white() + theme(legend.text = element_text(size=8)) + labs(x="Year Leader's Spell Ended", y="Tenure of Leader (not cumulative)", color="Method of Entry") ``` ] .pull-right[ ``` ## Warning in grid.Call.graphics(C_points, x$x, x$y, x$pch, x$size): semi- ## transparency is not supported on this device: reported only once per page ``` <!-- --> ] --- class: inverse background-image: url("img/BGGraph-alpha.svg") background-size: contain # Layers -- .pull-left[ Color aesthetics: - **color and/or fill** - Color colors the shape outline - Fill fills the shape interior ```r Archigos %>% *ggplot(., aes(x=endYear, y=tenureY, color=entry)) + geom_jitter(alpha=0.3) + scale_color_viridis_d() + theme(legend.text = element_text(size=8)) + labs(x="Year Leader's Spell Ended", y="Tenure of Leader (not cumulative)", color="Method of Entry") ``` ] .pull-right[ ``` ## Warning in grid.Call.graphics(C_points, x$x, x$y, x$pch, x$size): semi- ## transparency is not supported on this device: reported only once per page ``` <!-- --> ] --- class: background-image: url("img/JitterC.svg") background-size: contain -- # Layers .pull-left[ - **shape** ```r Archigos %>% ggplot(., aes(x=endYear, y=tenureY, color=entry, *shape=gender)) + geom_jitter(alpha=0.3) + scale_fill_viridis_d() + theme(legend.text = element_text(size=8)) + labs(x="Year Leader's Spell Ended", y="Tenure of Leader (not cumulative)", color="Method of Entry", shape="Gender") ``` ] .pull-right[ ``` ## Warning in grid.Call.graphics(C_points, x$x, x$y, x$pch, x$size): semi- ## transparency is not supported on this device: reported only once per page ``` <!-- --> ] --- background-image: url("img/LayerMF.svg") background-size: contain -- # Layers .pull-left[ **label** comes in two forms - geom_text() for text - geom_label for a text box<sup>1<sup> ```r Archigos %>% ggplot(., aes(x=endYear, y=tenureY, color=entry, label=gender)) + geom_label(alpha=0.2) + theme_economist_white() + theme(legend.text = element_text(size=7)) + labs(x="Year Leader's Spell Ended", y="Tenure of Leader (not cumulative)", color="Method of Entry", shape="Gender", title="Gender, Method of Entry, Tenure, and Year Spell Ends") ``` ] .pull-right[ <!-- --> ] .footnote[ [1] The excellent `ggrepel` package works with spatial tools to assist the management of placement and collisions of text. ] --- ## Layers What determines x and y? **Nothing.** Coordinates are arbitrary. Frequency determines the size.  --- # We can do that. Size is the number of previous times in office. <!-- --> --- class: inverse # A Grammar of Graphics - data - a mapping from data to aesthetics - a geometry for representing the aesthetic We deployed shapes, colors, text, size, color, fill, and alpha with one or two coordinates to create a fairly extensive set of visualisations of leaders. We also succeeded in displaying five distinct bits of data in one flat design. --- # Adding Further Information Requires Time A terrible pun. A map is a geometry. Color displays subprime population in the polygons. It changes through time.  --- class: inverse # That's Timely  --- class: inverse # But Seriously.... x is the ending year of the leader's spell with tenure mapped to y. We color entry by the key, fill exit by a key, label by gender, and make it evolve by entry time of the leader. Six dimensions. ``` ## Rendering [--------------------------------------------] at 6.1 fps ~ eta: 1m Rendering [--------------------------------------------] at 5.9 fps ~ eta: 1m Rendering [>---------------------------------------------] at 6 fps ~ eta: 1m Rendering [>-------------------------------------------] at 5.9 fps ~ eta: 1m Rendering [>-------------------------------------------] at 5.8 fps ~ eta: 1m Rendering [=>------------------------------------------] at 5.8 fps ~ eta: 1m Rendering [=>------------------------------------------] at 5.7 fps ~ eta: 1m Rendering [=>------------------------------------------] at 5.8 fps ~ eta: 1m Rendering [=>------------------------------------------] at 5.7 fps ~ eta: 1m Rendering [=>------------------------------------------] at 5.8 fps ~ eta: 1m Rendering [==>-----------------------------------------] at 5.6 fps ~ eta: 1m Rendering [==>-----------------------------------------] at 5.5 fps ~ eta: 1m Rendering [==>-----------------------------------------] at 5.6 fps ~ eta: 1m Rendering [===>----------------------------------------] at 5.6 fps ~ eta: 1m Rendering [====>---------------------------------------] at 5.6 fps ~ eta: 1m Rendering [====>---------------------------------------] at 5.7 fps ~ eta: 1m Rendering [====>---------------------------------------] at 5.6 fps ~ eta: 1m Rendering [====>---------------------------------------] at 5.7 fps ~ eta: 1m Rendering [=====>--------------------------------------] at 5.7 fps ~ eta: 1m Rendering [=====>--------------------------------------] at 5.6 fps ~ eta: 1m Rendering [======>-------------------------------------] at 5.6 fps ~ eta: 1m Rendering [=======>------------------------------------] at 5.6 fps ~ eta: 1m Rendering [========>-----------------------------------] at 5.6 fps ~ eta: 1m Rendering [=========>----------------------------------] at 5.6 fps ~ eta: 1m Rendering [=========>----------------------------------] at 5.5 fps ~ eta: 1m Rendering [==========>---------------------------------] at 5.5 fps ~ eta: 1m Rendering [===========>--------------------------------] at 5.5 fps ~ eta: 1m Rendering [============>-------------------------------] at 5.5 fps ~ eta: 1m Rendering [=============>------------------------------] at 5.5 fps ~ eta: 1m Rendering [=============>------------------------------] at 5.5 fps ~ eta: 50s Rendering [=============>------------------------------] at 5.5 fps ~ eta: 49s Rendering [==============>-----------------------------] at 5.5 fps ~ eta: 49s Rendering [==============>-----------------------------] at 5.5 fps ~ eta: 48s Rendering [==============>-----------------------------] at 5.4 fps ~ eta: 48s Rendering [===============>----------------------------] at 5.4 fps ~ eta: 48s Rendering [===============>----------------------------] at 5.4 fps ~ eta: 47s Rendering [===============>----------------------------] at 5.4 fps ~ eta: 46s Rendering [================>---------------------------] at 5.4 fps ~ eta: 46s Rendering [================>---------------------------] at 5.4 fps ~ eta: 45s Rendering [=================>--------------------------] at 5.4 fps ~ eta: 45s Rendering [=================>--------------------------] at 5.4 fps ~ eta: 44s Rendering [=================>--------------------------] at 5.4 fps ~ eta: 43s Rendering [==================>-------------------------] at 5.4 fps ~ eta: 43s Rendering [==================>-------------------------] at 5.3 fps ~ eta: 43s Rendering [==================>-------------------------] at 5.3 fps ~ eta: 42s Rendering [===================>------------------------] at 5.3 fps ~ eta: 42s Rendering [===================>------------------------] at 5.3 fps ~ eta: 41s Rendering [===================>------------------------] at 5.3 fps ~ eta: 40s Rendering [====================>-----------------------] at 5.3 fps ~ eta: 40s Rendering [====================>-----------------------] at 5.3 fps ~ eta: 39s Rendering [=====================>----------------------] at 5.3 fps ~ eta: 39s Rendering [=====================>----------------------] at 5.3 fps ~ eta: 38s Rendering [=====================>----------------------] at 5.2 fps ~ eta: 37s Rendering [======================>---------------------] at 5.2 fps ~ eta: 37s Rendering [======================>---------------------] at 5.2 fps ~ eta: 36s Rendering [=======================>--------------------] at 5.2 fps ~ eta: 36s Rendering [=======================>--------------------] at 5.2 fps ~ eta: 35s Rendering [========================>-------------------] at 5.1 fps ~ eta: 34s Rendering [========================>-------------------] at 5.1 fps ~ eta: 33s Rendering [=========================>------------------] at 5.1 fps ~ eta: 33s Rendering [=========================>------------------] at 5.1 fps ~ eta: 32s Rendering [==========================>-----------------] at 5.1 fps ~ eta: 31s Rendering [===========================>------------------] at 5 fps ~ eta: 31s Rendering [===========================>------------------] at 5 fps ~ eta: 30s Rendering [============================>-----------------] at 5 fps ~ eta: 30s Rendering [============================>-----------------] at 5 fps ~ eta: 29s Rendering [=============================>----------------] at 5 fps ~ eta: 29s Rendering [=============================>----------------] at 5 fps ~ eta: 28s Rendering [============================>---------------] at 4.9 fps ~ eta: 28s Rendering [============================>---------------] at 4.9 fps ~ eta: 27s Rendering [=============================>--------------] at 4.9 fps ~ eta: 27s Rendering [=============================>--------------] at 4.9 fps ~ eta: 26s Rendering [=============================>--------------] at 4.9 fps ~ eta: 25s Rendering [==============================>-------------] at 4.9 fps ~ eta: 25s Rendering [==============================>-------------] at 4.9 fps ~ eta: 24s Rendering [==============================>-------------] at 4.9 fps ~ eta: 23s Rendering [===============================>------------] at 4.9 fps ~ eta: 23s Rendering [===============================>------------] at 4.9 fps ~ eta: 22s Rendering [===============================>------------] at 4.9 fps ~ eta: 21s Rendering [================================>-----------] at 4.9 fps ~ eta: 21s Rendering [================================>-----------] at 4.9 fps ~ eta: 20s Rendering [=================================>----------] at 4.9 fps ~ eta: 19s Rendering [=================================>----------] at 4.9 fps ~ eta: 18s Rendering [==================================>---------] at 4.9 fps ~ eta: 18s Rendering [==================================>---------] at 4.9 fps ~ eta: 17s Rendering [==================================>---------] at 4.9 fps ~ eta: 16s Rendering [===================================>--------] at 4.9 fps ~ eta: 16s Rendering [===================================>--------] at 4.8 fps ~ eta: 15s Rendering [===================================>--------] at 4.8 fps ~ eta: 14s Rendering [====================================>-------] at 4.8 fps ~ eta: 14s Rendering [====================================>-------] at 4.8 fps ~ eta: 13s Rendering [====================================>-------] at 4.8 fps ~ eta: 12s Rendering [=====================================>------] at 4.8 fps ~ eta: 12s Rendering [=====================================>------] at 4.8 fps ~ eta: 11s Rendering [=====================================>------] at 4.8 fps ~ eta: 10s Rendering [======================================>-----] at 4.8 fps ~ eta: 10s Rendering [======================================>-----] at 4.8 fps ~ eta: 9s Rendering [=======================================>----] at 4.8 fps ~ eta: 8s Rendering [=======================================>----] at 4.8 fps ~ eta: 7s Rendering [========================================>---] at 4.8 fps ~ eta: 7s Rendering [========================================>---] at 4.8 fps ~ eta: 6s Rendering [========================================>---] at 4.8 fps ~ eta: 5s Rendering [========================================>---] at 4.7 fps ~ eta: 5s Rendering [=========================================>--] at 4.7 fps ~ eta: 5s Rendering [=========================================>--] at 4.7 fps ~ eta: 4s Rendering [=========================================>--] at 4.7 fps ~ eta: 3s Rendering [==========================================>-] at 4.7 fps ~ eta: 3s Rendering [==========================================>-] at 4.7 fps ~ eta: 2s Rendering [==========================================>-] at 4.7 fps ~ eta: 1s Rendering [===========================================>] at 4.7 fps ~ eta: 1s Rendering [===========================================>] at 4.7 fps ~ eta: 0s Rendering [============================================] at 4.7 fps ~ eta: 0s ``` ``` ## Frame 1 (0%) Frame 2 (0%) Frame 3 (0%) Frame 4 (1%) Frame 5 (1%) Frame 6 (1%) Frame 7 (1%) Frame 8 (2%) Frame 9 (2%) Frame 10 (2%) Frame 11 (2%) Frame 12 (3%) Frame 13 (3%) Frame 14 (3%) Frame 15 (3%) Frame 16 (4%) Frame 17 (4%) Frame 18 (4%) Frame 19 (4%) Frame 20 (5%) Frame 21 (5%) Frame 22 (5%) Frame 23 (5%) Frame 24 (6%) Frame 25 (6%) Frame 26 (6%) Frame 27 (6%) Frame 28 (7%) Frame 29 (7%) Frame 30 (7%) Frame 31 (7%) Frame 32 (8%) Frame 33 (8%) Frame 34 (8%) Frame 35 (8%) Frame 36 (9%) Frame 37 (9%) Frame 38 (9%) Frame 39 (9%) Frame 40 (10%) Frame 41 (10%) Frame 42 (10%) Frame 43 (10%) Frame 44 (11%) Frame 45 (11%) Frame 46 (11%) Frame 47 (11%) Frame 48 (12%) Frame 49 (12%) Frame 50 (12%) Frame 51 (12%) Frame 52 (13%) Frame 53 (13%) Frame 54 (13%) Frame 55 (13%) Frame 56 (14%) Frame 57 (14%) Frame 58 (14%) Frame 59 (14%) Frame 60 (15%) Frame 61 (15%) Frame 62 (15%) Frame 63 (15%) Frame 64 (16%) Frame 65 (16%) Frame 66 (16%) Frame 67 (16%) Frame 68 (17%) Frame 69 (17%) Frame 70 (17%) Frame 71 (17%) Frame 72 (18%) Frame 73 (18%) Frame 74 (18%) Frame 75 (18%) Frame 76 (19%) Frame 77 (19%) Frame 78 (19%) Frame 79 (19%) Frame 80 (20%) Frame 81 (20%) Frame 82 (20%) Frame 83 (20%) Frame 84 (21%) Frame 85 (21%) Frame 86 (21%) Frame 87 (21%) Frame 88 (22%) Frame 89 (22%) Frame 90 (22%) Frame 91 (22%) Frame 92 (23%) Frame 93 (23%) Frame 94 (23%) Frame 95 (23%) Frame 96 (24%) Frame 97 (24%) Frame 98 (24%) Frame 99 (24%) Frame 100 (25%) Frame 101 (25%) Frame 102 (25%) Frame 103 (25%) Frame 104 (26%) Frame 105 (26%) Frame 106 (26%) Frame 107 (26%) Frame 108 (27%) Frame 109 (27%) Frame 110 (27%) Frame 111 (27%) Frame 112 (28%) Frame 113 (28%) Frame 114 (28%) Frame 115 (28%) Frame 116 (29%) Frame 117 (29%) Frame 118 (29%) Frame 119 (29%) Frame 120 (30%) Frame 121 (30%) Frame 122 (30%) Frame 123 (30%) Frame 124 (31%) Frame 125 (31%) Frame 126 (31%) Frame 127 (31%) Frame 128 (32%) Frame 129 (32%) Frame 130 (32%) Frame 131 (32%) Frame 132 (33%) Frame 133 (33%) Frame 134 (33%) Frame 135 (33%) Frame 136 (34%) Frame 137 (34%) Frame 138 (34%) Frame 139 (34%) Frame 140 (35%) Frame 141 (35%) Frame 142 (35%) Frame 143 (35%) Frame 144 (36%) Frame 145 (36%) Frame 146 (36%) Frame 147 (36%) Frame 148 (37%) Frame 149 (37%) Frame 150 (37%) Frame 151 (37%) Frame 152 (38%) Frame 153 (38%) Frame 154 (38%) Frame 155 (38%) Frame 156 (39%) Frame 157 (39%) Frame 158 (39%) Frame 159 (39%) Frame 160 (40%) Frame 161 (40%) Frame 162 (40%) Frame 163 (40%) Frame 164 (41%) Frame 165 (41%) Frame 166 (41%) Frame 167 (41%) Frame 168 (42%) Frame 169 (42%) Frame 170 (42%) Frame 171 (42%) Frame 172 (43%) Frame 173 (43%) Frame 174 (43%) Frame 175 (43%) Frame 176 (44%) Frame 177 (44%) Frame 178 (44%) Frame 179 (44%) Frame 180 (45%) Frame 181 (45%) Frame 182 (45%) Frame 183 (45%) Frame 184 (46%) Frame 185 (46%) Frame 186 (46%) Frame 187 (46%) Frame 188 (47%) Frame 189 (47%) Frame 190 (47%) Frame 191 (47%) Frame 192 (48%) Frame 193 (48%) Frame 194 (48%) Frame 195 (48%) Frame 196 (49%) Frame 197 (49%) Frame 198 (49%) Frame 199 (49%) Frame 200 (50%) Frame 201 (50%) Frame 202 (50%) Frame 203 (50%) Frame 204 (51%) Frame 205 (51%) Frame 206 (51%) Frame 207 (51%) Frame 208 (52%) Frame 209 (52%) Frame 210 (52%) Frame 211 (52%) Frame 212 (53%) Frame 213 (53%) Frame 214 (53%) Frame 215 (53%) Frame 216 (54%) Frame 217 (54%) Frame 218 (54%) Frame 219 (54%) Frame 220 (55%) Frame 221 (55%) Frame 222 (55%) Frame 223 (55%) Frame 224 (56%) Frame 225 (56%) Frame 226 (56%) Frame 227 (56%) Frame 228 (57%) Frame 229 (57%) Frame 230 (57%) Frame 231 (57%) Frame 232 (58%) Frame 233 (58%) Frame 234 (58%) Frame 235 (58%) Frame 236 (59%) Frame 237 (59%) Frame 238 (59%) Frame 239 (59%) Frame 240 (60%) Frame 241 (60%) Frame 242 (60%) Frame 243 (60%) Frame 244 (61%) Frame 245 (61%) Frame 246 (61%) Frame 247 (61%) Frame 248 (62%) Frame 249 (62%) Frame 250 (62%) Frame 251 (62%) Frame 252 (63%) Frame 253 (63%) Frame 254 (63%) Frame 255 (63%) Frame 256 (64%) Frame 257 (64%) Frame 258 (64%) Frame 259 (64%) Frame 260 (65%) Frame 261 (65%) Frame 262 (65%) Frame 263 (65%) Frame 264 (66%) Frame 265 (66%) Frame 266 (66%) Frame 267 (66%) Frame 268 (67%) Frame 269 (67%) Frame 270 (67%) Frame 271 (67%) Frame 272 (68%) Frame 273 (68%) Frame 274 (68%) Frame 275 (68%) Frame 276 (69%) Frame 277 (69%) Frame 278 (69%) Frame 279 (69%) Frame 280 (70%) Frame 281 (70%) Frame 282 (70%) Frame 283 (70%) Frame 284 (71%) Frame 285 (71%) Frame 286 (71%) Frame 287 (71%) Frame 288 (72%) Frame 289 (72%) Frame 290 (72%) Frame 291 (72%) Frame 292 (73%) Frame 293 (73%) Frame 294 (73%) Frame 295 (73%) Frame 296 (74%) Frame 297 (74%) Frame 298 (74%) Frame 299 (74%) Frame 300 (75%) Frame 301 (75%) Frame 302 (75%) Frame 303 (75%) Frame 304 (76%) Frame 305 (76%) Frame 306 (76%) Frame 307 (76%) Frame 308 (77%) Frame 309 (77%) Frame 310 (77%) Frame 311 (77%) Frame 312 (78%) Frame 313 (78%) Frame 314 (78%) Frame 315 (78%) Frame 316 (79%) Frame 317 (79%) Frame 318 (79%) Frame 319 (79%) Frame 320 (80%) Frame 321 (80%) Frame 322 (80%) Frame 323 (80%) Frame 324 (81%) Frame 325 (81%) Frame 326 (81%) Frame 327 (81%) Frame 328 (82%) Frame 329 (82%) Frame 330 (82%) Frame 331 (82%) Frame 332 (83%) Frame 333 (83%) Frame 334 (83%) Frame 335 (83%) Frame 336 (84%) Frame 337 (84%) Frame 338 (84%) Frame 339 (84%) Frame 340 (85%) Frame 341 (85%) Frame 342 (85%) Frame 343 (85%) Frame 344 (86%) Frame 345 (86%) Frame 346 (86%) Frame 347 (86%) Frame 348 (87%) Frame 349 (87%) Frame 350 (87%) Frame 351 (87%) Frame 352 (88%) Frame 353 (88%) Frame 354 (88%) Frame 355 (88%) Frame 356 (89%) Frame 357 (89%) Frame 358 (89%) Frame 359 (89%) Frame 360 (90%) Frame 361 (90%) Frame 362 (90%) Frame 363 (90%) Frame 364 (91%) Frame 365 (91%) Frame 366 (91%) Frame 367 (91%) Frame 368 (92%) Frame 369 (92%) Frame 370 (92%) Frame 371 (92%) Frame 372 (93%) Frame 373 (93%) Frame 374 (93%) Frame 375 (93%) Frame 376 (94%) Frame 377 (94%) Frame 378 (94%) Frame 379 (94%) Frame 380 (95%) Frame 381 (95%) Frame 382 (95%) Frame 383 (95%) Frame 384 (96%) Frame 385 (96%) Frame 386 (96%) Frame 387 (96%) Frame 388 (97%) Frame 389 (97%) Frame 390 (97%) Frame 391 (97%) Frame 392 (98%) Frame 393 (98%) Frame 394 (98%) Frame 395 (98%) Frame 396 (99%) Frame 397 (99%) Frame 398 (99%) Frame 399 (99%) Frame 400 (100%) ## Finalizing encoding... done! ``` <!-- --> --- class: inverse # The Takeaway Visualisations are one or two [discrete or quantitative] coordinates. We map data to aesthetics -- color and fill, transparency, label and text, shape, and size -- to show points, lines [connections], or areas. `ggplot` writes graphics that mirror this exact language: data, aes(), and `geom_` --- class: bottom background-image: url("img/PicAnim.gif") background-position: 65% 8% # Thanks! Slides created via the R package [**xaringan**](https://github.com/yihui/xaringan). Theme borrowed from `@apreshill` and her `r-ladies` theme.