{"id":853,"date":"2015-04-07T11:35:08","date_gmt":"2015-04-07T18:35:08","guid":{"rendered":"http:\/\/homepages.uc.edu\/~yaozo\/wordpress\/?p=853"},"modified":"2015-04-07T11:35:08","modified_gmt":"2015-04-07T18:35:08","slug":"ggplot2-cheatsheet-from-r-for-public-health","status":"publish","type":"post","link":"https:\/\/zhuoyao.net\/index.php\/2015\/04\/07\/ggplot2-cheatsheet-from-r-for-public-health\/","title":{"rendered":"ggplot2 Cheatsheet from R for Public Health"},"content":{"rendered":"
\n
\n

R for Public Health<\/a><\/p>\n

\n
library<\/span>(<\/span>ggplot2)<\/span>\nlibrary<\/span>(<\/span>gridExtra)<\/span>\nmtc <-<\/span> mtcars\nhead<\/span>(<\/span>mtc)<\/span>\n<\/code><\/pre>\n<\/div>\n
\n
                   mpg cyl disp  hp drat    wt  qsec vs am gear carb\nMazda RX4         21.0   6  160 110 3.90 2.620 16.46  0  1    4    4\nMazda RX4 Wag     21.0   6  160 110 3.90 2.875 17.02  0  1    4    4\nDatsun 710        22.8   4  108  93 3.85 2.320 18.61  1  1    4    1\nHornet 4 Drive    21.4   6  258 110 3.08 3.215 19.44  1  0    3    1\nHornet Sportabout 18.7   8  360 175 3.15 3.440 17.02  0  0    3    2\nValiant           18.1   6  225 105 2.76 3.460 20.22  1  0    3    1\n<\/code><\/pre>\n<\/div>\n
Scatterplots<\/h4>\n
Basic scatterplot<\/h5>\n
\n
p1 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> hp,<\/span> y =<\/span> mpg))<\/span>\n# Print plot with default points<\/span>\np1 +<\/span> geom_point()<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Change color of points<\/h5>\n
\n
p2 <-<\/span> p1 +<\/span> geom_point(<\/span>color =<\/span> \"red\"<\/span>)<\/span>  #set one color for all points<\/span>\np3 <-<\/span> p1 +<\/span> geom_point(<\/span>aes(<\/span>color =<\/span> wt))<\/span>  #set color scale by a continuous variable<\/span>\np4 <-<\/span> p1 +<\/span> geom_point(<\/span>aes(<\/span>color =<\/span> factor<\/span>(<\/span>am)))<\/span>  #set color scale by a factor variable<\/span>\ngrid.arrange(<\/span>p2,<\/span> p3,<\/span> p4,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
\n
# Change default colors in color scale<\/span>\np1 +<\/span> geom_point(<\/span>aes(<\/span>color =<\/span> factor<\/span>(<\/span>am)))<\/span> +<\/span> scale_color_manual(<\/span>values =<\/span> c<\/span>(<\/span>\"orange\"<\/span>,<\/span> \n    \"purple\"<\/span>))<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Change shape or size of points<\/h5>\n
\n
p2 <-<\/span> p1 +<\/span> geom_point(<\/span>size =<\/span> 5<\/span>)<\/span>  #increase all points to size 5<\/span>\np3 <-<\/span> p1 +<\/span> geom_point(<\/span>aes(<\/span>size =<\/span> wt))<\/span>  #set point size by continuous variable<\/span>\np4 <-<\/span> p1 +<\/span> geom_point(<\/span>aes(<\/span>shape =<\/span> factor<\/span>(<\/span>am)))<\/span>  #set point shape by factor variable    <\/span>\ngrid.arrange(<\/span>p2,<\/span> p3,<\/span> p4,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
\n
# change the default shapes<\/span>\np1 +<\/span> geom_point(<\/span>aes(<\/span>shape =<\/span> factor<\/span>(<\/span>am)))<\/span> +<\/span> scale_shape_manual(<\/span>values =<\/span> c<\/span>(<\/span>0<\/span>,<\/span> \n    2<\/span>))<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Add lines to scatterplot<\/h5>\n
\n
p2 <-<\/span> p1 +<\/span> geom_point(<\/span>color =<\/span> \"blue\"<\/span>)<\/span> +<\/span> geom_line()<\/span>  #connect points with line<\/span>\np3 <-<\/span> p1 +<\/span> geom_point(<\/span>color =<\/span> \"red\"<\/span>)<\/span> +<\/span> geom_smooth(<\/span>method =<\/span> \"lm\"<\/span>,<\/span> se =<\/span> TRUE<\/span>)<\/span>  #add regression line<\/span>\np4 <-<\/span> p1 +<\/span> geom_point()<\/span> +<\/span> geom_vline(<\/span>xintercept =<\/span> 100<\/span>,<\/span> color =<\/span> \"red\"<\/span>)<\/span>  #add vertical line<\/span>\ngrid.arrange(<\/span>p2,<\/span> p3,<\/span> p4,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
\n
# take out the points, and just create a line plot, and change size and<\/span>\n# color as before<\/span>\nggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> wt,<\/span> y =<\/span> qsec))<\/span> +<\/span> geom_line(<\/span>size =<\/span> 2<\/span>,<\/span> aes(<\/span>color =<\/span> factor<\/span>(<\/span>vs)))<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Change axis labels<\/h5>\n
\n
p2 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> hp,<\/span> y =<\/span> mpg))<\/span> +<\/span> geom_point()<\/span>\n\n# label all axes at once<\/span>\np3 <-<\/span> p2 +<\/span> labs(<\/span>x =<\/span> \"Horsepower\"<\/span>,<\/span> y =<\/span> \"Miles per Gallon\"<\/span>)<\/span>\n\n# label and change font size<\/span>\np4 <-<\/span> p2 +<\/span> theme(<\/span>axis.title.x =<\/span> element_text(<\/span>face =<\/span> \"bold\"<\/span>,<\/span> size =<\/span> 20<\/span>))<\/span> +<\/span> labs(<\/span>x =<\/span> \"Horsepower\"<\/span>)<\/span>\n\n# adjust axis limits and breaks<\/span>\np5 <-<\/span> p2 +<\/span> scale_x_continuous(<\/span>\"Horsepower\"<\/span>,<\/span> limits =<\/span> c<\/span>(<\/span>0<\/span>,<\/span> 400<\/span>),<\/span> breaks =<\/span> seq<\/span>(<\/span>0<\/span>,<\/span> \n    400<\/span>,<\/span> 50<\/span>))<\/span>\n\ngrid.arrange(<\/span>p3,<\/span> p4,<\/span> p5,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Change legend options<\/h5>\n
\n
g1<-<\/span>ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> hp,<\/span> y =<\/span> mpg))<\/span> +<\/span> geom_point(<\/span>aes(<\/span>color=<\/span>factor<\/span>(<\/span>vs)))<\/span>\n#move legend inside<\/span>\ng2 <-<\/span> g1 +<\/span> theme(<\/span>legend.position=<\/span>c<\/span>(<\/span>1<\/span>,<\/span>1<\/span>),<\/span>legend.justification=<\/span>c<\/span>(<\/span>1<\/span>,<\/span>1<\/span>))<\/span> \n#move legend bottom <\/span>\ng3 <-<\/span> g1 +<\/span> theme(<\/span>legend.position =<\/span> \"bottom\"<\/span>)<\/span>    \n#change labels<\/span>\ng4 <-<\/span> g1 +<\/span> scale_color_discrete(<\/span>name =<\/span>\"Engine\"<\/span>,<\/span> labels=<\/span>c<\/span>(<\/span>\"V-engine\"<\/span>,<\/span> \"Straight engine\"<\/span>))<\/span> \ngrid.arrange(<\/span>g2,<\/span> g3,<\/span> g4,<\/span> nrow=<\/span>1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
\n
g5<-<\/span>ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> hp,<\/span> y =<\/span> mpg))<\/span> +<\/span> geom_point(<\/span>size=<\/span>2<\/span>,<\/span> aes(<\/span>color =<\/span> wt))<\/span>\ng5 +<\/span> scale_color_continuous(<\/span>name=<\/span>\"Weight\"<\/span>,<\/span> #name of legend<\/span>\n                            breaks =<\/span> with<\/span>(<\/span>mtc,<\/span> c<\/span>(<\/span>min<\/span>(<\/span>wt),<\/span> mean<\/span>(<\/span>wt),<\/span> max<\/span>(<\/span>wt))),<\/span> #choose breaks of variable<\/span>\n                            labels =<\/span> c<\/span>(<\/span>\"Light\"<\/span>,<\/span> \"Medium\"<\/span>,<\/span> \"Heavy\"<\/span>),<\/span> #label<\/span>\n                            low =<\/span> \"pink\"<\/span>,<\/span>  #color of lowest value<\/span>\n                            high =<\/span> \"red\"<\/span>  #color of highest value<\/span>\n                            )<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Change background color and style<\/h5>\n
\n
g2 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> hp,<\/span> y =<\/span> mpg))<\/span> +<\/span> geom_point()<\/span>\n\n# Completely clear all lines except axis lines and make background white<\/span>\nt1 <-<\/span> theme(<\/span>plot.background =<\/span> element_blank(),<\/span> panel.grid.major =<\/span> element_blank(),<\/span> \n    panel.grid.minor =<\/span> element_blank(),<\/span> panel.border =<\/span> element_blank(),<\/span> panel.background =<\/span> element_blank(),<\/span> \n    axis.line =<\/span> element_line(<\/span>size =<\/span> 0.4<\/span>))<\/span>\n\n# Use theme to change axis label style<\/span>\nt2 <-<\/span> theme(<\/span>axis.title.x =<\/span> element_text(<\/span>face =<\/span> \"bold\"<\/span>,<\/span> color =<\/span> \"black\"<\/span>,<\/span> size =<\/span> 10<\/span>),<\/span> \n    axis.title.y =<\/span> element_text(<\/span>face =<\/span> \"bold\"<\/span>,<\/span> color =<\/span> \"black\"<\/span>,<\/span> size =<\/span> 10<\/span>),<\/span> \n    plot.title =<\/span> element_text(<\/span>face =<\/span> \"bold\"<\/span>,<\/span> color =<\/span> \"black\"<\/span>,<\/span> size =<\/span> 12<\/span>))<\/span>\n\ng3 <-<\/span> g2 +<\/span> t1\ng4 <-<\/span> g2 +<\/span> theme_bw()<\/span>\ng5 <-<\/span> g2 +<\/span> theme_bw()<\/span> +<\/span> t2 +<\/span> labs(<\/span>x =<\/span> \"Horsepower\"<\/span>,<\/span> y =<\/span> \"Miles per Gallon\"<\/span>,<\/span> \n    title =<\/span> \"MPG vs Horsepower\"<\/span>)<\/span>\ngrid.arrange(<\/span>g2,<\/span> g3,<\/span> g4,<\/span> g5,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
a nice graph using a combination of options<\/h5>\n
\n
g2 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> hp,<\/span> y =<\/span> mpg))<\/span> +<\/span> geom_point(<\/span>size =<\/span> 2<\/span>,<\/span> aes(<\/span>color =<\/span> factor<\/span>(<\/span>vs),<\/span> \n    shape =<\/span> factor<\/span>(<\/span>vs)))<\/span> +<\/span> geom_smooth(<\/span>aes(<\/span>color =<\/span> factor<\/span>(<\/span>vs)),<\/span> method =<\/span> \"lm\"<\/span>,<\/span> \n    se =<\/span> TRUE<\/span>)<\/span> +<\/span> scale_color_manual(<\/span>name =<\/span> \"Engine\"<\/span>,<\/span> labels =<\/span> c<\/span>(<\/span>\"V-engine\"<\/span>,<\/span> \n    \"Straight engine\"<\/span>),<\/span> values =<\/span> c<\/span>(<\/span>\"red\"<\/span>,<\/span> \"blue\"<\/span>))<\/span> +<\/span> scale_shape_manual(<\/span>name =<\/span> \"Engine\"<\/span>,<\/span> \n    labels =<\/span> c<\/span>(<\/span>\"V-engine\"<\/span>,<\/span> \"Straight engine\"<\/span>),<\/span> values =<\/span> c<\/span>(<\/span>0<\/span>,<\/span> 2<\/span>))<\/span> +<\/span> theme_bw()<\/span> +<\/span> \n    theme(<\/span>axis.title.x =<\/span> element_text(<\/span>face =<\/span> \"bold\"<\/span>,<\/span> color =<\/span> \"black\"<\/span>,<\/span> size =<\/span> 12<\/span>),<\/span> \n        axis.title.y =<\/span> element_text(<\/span>face =<\/span> \"bold\"<\/span>,<\/span> color =<\/span> \"black\"<\/span>,<\/span> size =<\/span> 12<\/span>),<\/span> \n        plot.title =<\/span> element_text(<\/span>face =<\/span> \"bold\"<\/span>,<\/span> color =<\/span> \"black\"<\/span>,<\/span> size =<\/span> 12<\/span>),<\/span> \n        legend.position =<\/span> c<\/span>(<\/span>1<\/span>,<\/span> 1<\/span>),<\/span> legend.justification =<\/span> c<\/span>(<\/span>1<\/span>,<\/span> 1<\/span>))<\/span> +<\/span> labs(<\/span>x =<\/span> \"Horsepower\"<\/span>,<\/span> \n    y =<\/span> \"Miles per Gallon\"<\/span>,<\/span> title =<\/span> \"Linear Regression (95% CI) of MPG vs Horsepower by Engine type\"<\/span>)<\/span>\n\ng2\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Barplots<\/h4>\n
Basic barplot<\/h5>\n
\n
ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear)))<\/span> +<\/span> geom_bar(<\/span>stat =<\/span> \"bin\"<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Horizontal bars, colors, width of bars<\/h5>\n
\n
# 1. horizontal bars<\/span>\np1 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> wt))<\/span> +<\/span> stat_summary(<\/span>fun.y =<\/span> mean<\/span>,<\/span> \n    geom =<\/span> \"bar\"<\/span>)<\/span> +<\/span> coord_flip()<\/span>\n\n# 2. change colors of bars<\/span>\np2 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> wt,<\/span> fill =<\/span> factor<\/span>(<\/span>gear)))<\/span> +<\/span> stat_summary(<\/span>fun.y =<\/span> mean<\/span>,<\/span> \n    geom =<\/span> \"bar\"<\/span>)<\/span> +<\/span> scale_fill_manual(<\/span>values =<\/span> c<\/span>(<\/span>\"purple\"<\/span>,<\/span> \"blue\"<\/span>,<\/span> \"darkgreen\"<\/span>))<\/span>\n\n# 3. change width of bars<\/span>\np3 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> wt))<\/span> +<\/span> stat_summary(<\/span>fun.y =<\/span> mean<\/span>,<\/span> \n    geom =<\/span> \"bar\"<\/span>,<\/span> aes(<\/span>width =<\/span> 0.5<\/span>))<\/span>\n\ngrid.arrange(<\/span>p1,<\/span> p2,<\/span> p3,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Split and color by another variable<\/h5>\n
\n
# 1. next to each other<\/span>\np1 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> wt,<\/span> fill =<\/span> factor<\/span>(<\/span>vs)),<\/span> color =<\/span> factor<\/span>(<\/span>vs))<\/span> +<\/span> \n    stat_summary(<\/span>fun.y =<\/span> mean<\/span>,<\/span> position =<\/span> position_dodge(),<\/span> geom =<\/span> \"bar\"<\/span>)<\/span>\n\n# 2. stacked<\/span>\np2 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> wt,<\/span> fill =<\/span> factor<\/span>(<\/span>vs)),<\/span> color =<\/span> factor<\/span>(<\/span>vs))<\/span> +<\/span> \n    stat_summary(<\/span>fun.y =<\/span> mean<\/span>,<\/span> position =<\/span> \"stack\"<\/span>,<\/span> geom =<\/span> \"bar\"<\/span>)<\/span>\n\n# 3. with facets<\/span>\np3 <-<\/span> ggplot(<\/span>mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> wt,<\/span> fill =<\/span> factor<\/span>(<\/span>vs)),<\/span> color =<\/span> factor<\/span>(<\/span>vs))<\/span> +<\/span> \n    stat_summary(<\/span>fun.y =<\/span> mean<\/span>,<\/span> geom =<\/span> \"bar\"<\/span>)<\/span> +<\/span> facet_wrap(<\/span>~<\/span>vs)<\/span>\n\ngrid.arrange(<\/span>p1,<\/span> p2,<\/span> p3,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Add text to the bars, label axes, and label legend<\/h5>\n
\n
ag.mtc <-<\/span> aggregate(<\/span>mtc$<\/span>wt,<\/span> by =<\/span> list<\/span>(<\/span>mtc$<\/span>gear,<\/span> mtc$<\/span>vs),<\/span> FUN =<\/span> mean<\/span>)<\/span>\ncolnames<\/span>(<\/span>ag.mtc)<\/span> <-<\/span> c<\/span>(<\/span>\"gear\"<\/span>,<\/span> \"vs\"<\/span>,<\/span> \"meanwt\"<\/span>)<\/span>\nag.mtc\n<\/code><\/pre>\n<\/div>\n
\n
  gear vs meanwt\n1    3  0  4.104\n2    4  0  2.748\n3    5  0  2.913\n4    3  1  3.047\n5    4  1  2.591\n6    5  1  1.513\n<\/code><\/pre>\n<\/div>\n
\n
# 1. basic<\/span>\ng1 <-<\/span> ggplot(<\/span>ag.mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> meanwt,<\/span> fill =<\/span> factor<\/span>(<\/span>vs),<\/span> color =<\/span> factor<\/span>(<\/span>vs)))<\/span> +<\/span> \n    geom_bar(<\/span>stat =<\/span> \"identity\"<\/span>,<\/span> position =<\/span> position_dodge())<\/span> +<\/span> geom_text(<\/span>aes(<\/span>y =<\/span> meanwt,<\/span> \n    ymax =<\/span> meanwt,<\/span> label =<\/span> meanwt),<\/span> position =<\/span> position_dodge(<\/span>width =<\/span> 0.9<\/span>),<\/span> \n    vjust =<\/span> -0.5<\/span>)<\/span>\n\n# 2. fixing the yaxis problem, changing the color of text, legend labels,<\/span>\n# and rounding to 2 decimals<\/span>\ng2 <-<\/span> ggplot(<\/span>ag.mtc,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> meanwt,<\/span> fill =<\/span> factor<\/span>(<\/span>vs)))<\/span> +<\/span> \n    geom_bar(<\/span>stat =<\/span> \"identity\"<\/span>,<\/span> position =<\/span> position_dodge())<\/span> +<\/span> geom_text(<\/span>aes(<\/span>y =<\/span> meanwt,<\/span> \n    ymax =<\/span> meanwt,<\/span> label =<\/span> round<\/span>(<\/span>meanwt,<\/span> 2<\/span>)),<\/span> position =<\/span> position_dodge(<\/span>width =<\/span> 0.9<\/span>),<\/span> \n    vjust =<\/span> -0.5<\/span>,<\/span> color =<\/span> \"black\"<\/span>)<\/span> +<\/span> scale_y_continuous(<\/span>\"Mean Weight\"<\/span>,<\/span> limits =<\/span> c<\/span>(<\/span>0<\/span>,<\/span> \n    4.5<\/span>),<\/span> breaks =<\/span> seq<\/span>(<\/span>0<\/span>,<\/span> 4.5<\/span>,<\/span> 0.5<\/span>))<\/span> +<\/span> scale_x_discrete(<\/span>\"Number of Gears\"<\/span>)<\/span> +<\/span> \n    scale_fill_discrete(<\/span>name =<\/span> \"Engine\"<\/span>,<\/span> labels =<\/span> c<\/span>(<\/span>\"V-engine\"<\/span>,<\/span> \"Straight engine\"<\/span>))<\/span>\n\ngrid.arrange(<\/span>g1,<\/span> g2,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Add error bars<\/h5>\n
\n
summary.mtc2 <-<\/span> data.frame<\/span>(<\/span>gear =<\/span> levels<\/span>(<\/span>as.factor<\/span>(<\/span>mtc$<\/span>gear)),<\/span> meanwt =<\/span> tapply<\/span>(<\/span>mtc$<\/span>wt,<\/span> \n    mtc$<\/span>gear,<\/span> mean<\/span>),<\/span> sd =<\/span> tapply<\/span>(<\/span>mtc$<\/span>wt,<\/span> mtc$<\/span>gear,<\/span> sd))<\/span>\nsummary.mtc2\n<\/code><\/pre>\n<\/div>\n
\n
  gear meanwt     sd\n3    3  3.893 0.8330\n4    4  2.617 0.6327\n5    5  2.633 0.8189\n<\/code><\/pre>\n<\/div>\n
\n
ggplot(<\/span>summary.mtc2,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>gear),<\/span> y =<\/span> meanwt))<\/span> +<\/span> geom_bar(<\/span>stat =<\/span> \"identity\"<\/span>,<\/span> \n    position =<\/span> \"dodge\"<\/span>,<\/span> fill =<\/span> \"lightblue\"<\/span>)<\/span> +<\/span> geom_errorbar(<\/span>aes(<\/span>ymin =<\/span> meanwt -<\/span> \n    sd,<\/span> ymax =<\/span> meanwt +<\/span> sd),<\/span> width =<\/span> 0.3<\/span>,<\/span> color =<\/span> \"darkblue\"<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Add best fit line<\/h5>\n
\n
# summarize data<\/span>\nsummary.mtc3 <-<\/span> data.frame<\/span>(<\/span>hp =<\/span> levels<\/span>(<\/span>as.factor<\/span>(<\/span>mtc$<\/span>hp)),<\/span> meanmpg =<\/span> tapply<\/span>(<\/span>mtc$<\/span>mpg,<\/span> \n    mtc$<\/span>hp,<\/span> mean<\/span>))<\/span>\n\n# run a model<\/span>\nl <-<\/span> summary<\/span>(<\/span>lm(<\/span>meanmpg ~<\/span> as.numeric<\/span>(<\/span>hp),<\/span> data =<\/span> summary.mtc3))<\/span>\n\n# manually entering the intercept and slope<\/span>\nf1 <-<\/span> ggplot(<\/span>summary.mtc3,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>hp),<\/span> y =<\/span> meanmpg))<\/span> +<\/span> geom_bar(<\/span>stat =<\/span> \"identity\"<\/span>,<\/span> \n    fill =<\/span> \"darkblue\"<\/span>)<\/span> +<\/span> geom_abline(<\/span>aes(<\/span>intercept =<\/span> l$<\/span>coef[<\/span>1<\/span>,<\/span> 1<\/span>],<\/span> slope =<\/span> l$<\/span>coef[<\/span>2<\/span>,<\/span> \n    1<\/span>]),<\/span> color =<\/span> \"red\"<\/span>,<\/span> size =<\/span> 1.5<\/span>)<\/span>\n\n# using stat_smooth to fit the line for you<\/span>\nf2 <-<\/span> ggplot(<\/span>summary.mtc3,<\/span> aes(<\/span>x =<\/span> factor<\/span>(<\/span>hp),<\/span> y =<\/span> meanmpg))<\/span> +<\/span> geom_bar(<\/span>stat =<\/span> \"identity\"<\/span>,<\/span> \n    fill =<\/span> \"darkblue\"<\/span>)<\/span> +<\/span> stat_smooth(<\/span>aes(<\/span>group =<\/span> 1<\/span>),<\/span> method =<\/span> \"lm\"<\/span>,<\/span> se =<\/span> FALSE<\/span>,<\/span> \n    color =<\/span> \"orange\"<\/span>,<\/span> size =<\/span> 1.5<\/span>)<\/span>\n\ngrid.arrange(<\/span>f1,<\/span> f2,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Histograms<\/h4>\n
\n
set.seed<\/span>(<\/span>999<\/span>)<\/span>\nxvar <-<\/span> c<\/span>(<\/span>rnorm(<\/span>1500<\/span>,<\/span> mean =<\/span> -1<\/span>),<\/span> rnorm(<\/span>1500<\/span>,<\/span> mean =<\/span> 1.5<\/span>))<\/span>\nyvar <-<\/span> c<\/span>(<\/span>rnorm(<\/span>1500<\/span>,<\/span> mean =<\/span> 1<\/span>),<\/span> rnorm(<\/span>1500<\/span>,<\/span> mean =<\/span> 1.5<\/span>))<\/span>\nzvar <-<\/span> as.factor<\/span>(<\/span>c<\/span>(<\/span>rep<\/span>(<\/span>1<\/span>,<\/span> 1500<\/span>),<\/span> rep<\/span>(<\/span>2<\/span>,<\/span> 1500<\/span>)))<\/span>\nxy <-<\/span> data.frame<\/span>(<\/span>xvar,<\/span> yvar,<\/span> zvar)<\/span>\n<\/code><\/pre>\n<\/div>\n
\n
# counts on y-axis<\/span>\ng1 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>xvar))<\/span> +<\/span> geom_histogram()<\/span>  #horribly ugly default<\/span>\ng2 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>xvar))<\/span> +<\/span> geom_histogram(<\/span>binwidth =<\/span> 1<\/span>)<\/span>  #change binwidth<\/span>\ng3 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>xvar))<\/span> +<\/span> geom_histogram(<\/span>fill =<\/span> NA<\/span>,<\/span> color =<\/span> \"black\"<\/span>)<\/span> +<\/span> theme_bw()<\/span>  #nicer looking<\/span>\n\n# density on y-axis<\/span>\ng4 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>x =<\/span> xvar))<\/span> +<\/span> geom_histogram(<\/span>aes(<\/span>y =<\/span> ..<\/span>density..),<\/span> color =<\/span> \"black\"<\/span>,<\/span> \n    fill =<\/span> NA<\/span>)<\/span> +<\/span> theme_bw()<\/span>\n\ngrid.arrange(<\/span>g1,<\/span> g2,<\/span> g3,<\/span> g4,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Density plots<\/h4>\n
\n
# basic density<\/span>\np1 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>xvar))<\/span> +<\/span> geom_density()<\/span>\n\n# histogram with density line overlaid<\/span>\np2 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>x =<\/span> xvar))<\/span> +<\/span> geom_histogram(<\/span>aes(<\/span>y =<\/span> ..<\/span>density..),<\/span> color =<\/span> \"black\"<\/span>,<\/span> \n    fill =<\/span> NA<\/span>)<\/span> +<\/span> geom_density(<\/span>color =<\/span> \"blue\"<\/span>)<\/span>\n\n# split and color by third variable, alpha fades the color a bit<\/span>\np3 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>xvar,<\/span> fill =<\/span> zvar))<\/span> +<\/span> geom_density(<\/span>alpha =<\/span> 0.2<\/span>)<\/span>\n\ngrid.arrange(<\/span>p1,<\/span> p2,<\/span> p3,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Boxplots<\/h4>\n
\n
# boxplot<\/span>\nb1 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>zvar,<\/span> xvar))<\/span> +<\/span> geom_boxplot(<\/span>aes(<\/span>fill =<\/span> zvar))<\/span> +<\/span> theme(<\/span>legend.position =<\/span> \"none\"<\/span>)<\/span>\n\n# jitter plot<\/span>\nb2 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>zvar,<\/span> xvar))<\/span> +<\/span> geom_jitter(<\/span>alpha =<\/span> I<\/span>(<\/span>1<\/span>\/<\/span>4<\/span>),<\/span> aes(<\/span>color =<\/span> zvar))<\/span> +<\/span> \n    theme(<\/span>legend.position =<\/span> \"none\"<\/span>)<\/span>\n\n# violin plot<\/span>\nb3 <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>x =<\/span> xvar))<\/span> +<\/span> stat_density(<\/span>aes(<\/span>ymax =<\/span> ..<\/span>density..,<\/span> ymin =<\/span> -..<\/span>density..,<\/span> \n    fill =<\/span> zvar,<\/span> color =<\/span> zvar),<\/span> geom =<\/span> \"ribbon\"<\/span>,<\/span> position =<\/span> \"identity\"<\/span>)<\/span> +<\/span> facet_grid(<\/span>.<\/span> ~<\/span> \n    zvar)<\/span> +<\/span> coord_flip()<\/span> +<\/span> theme(<\/span>legend.position =<\/span> \"none\"<\/span>)<\/span>\n\ngrid.arrange(<\/span>b1,<\/span> b2,<\/span> b3,<\/span> nrow =<\/span> 1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Putting multiple plots together<\/h4>\n
\n
# rug plot<\/span>\nggplot(<\/span>xy,<\/span> aes(<\/span>xvar,<\/span> yvar))<\/span> +<\/span> geom_point()<\/span> +<\/span> geom_rug(<\/span>col =<\/span> \"darkred\"<\/span>,<\/span> alpha =<\/span> 0.1<\/span>)<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
\n
# placeholder plot - prints nothing at all<\/span>\nempty <-<\/span> ggplot()<\/span> +<\/span> geom_point(<\/span>aes(<\/span>1<\/span>,<\/span> 1<\/span>),<\/span> colour =<\/span> \"white\"<\/span>)<\/span> +<\/span> theme(<\/span>plot.background =<\/span> element_blank(),<\/span> \n    panel.grid.major =<\/span> element_blank(),<\/span> panel.grid.minor =<\/span> element_blank(),<\/span> \n    panel.border =<\/span> element_blank(),<\/span> panel.background =<\/span> element_blank(),<\/span> axis.title.x =<\/span> element_blank(),<\/span> \n    axis.title.y =<\/span> element_blank(),<\/span> axis.text.x =<\/span> element_blank(),<\/span> axis.text.y =<\/span> element_blank(),<\/span> \n    axis.ticks =<\/span> element_blank())<\/span>\n\n# scatterplot of x and y variables<\/span>\nscatter <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>xvar,<\/span> yvar))<\/span> +<\/span> geom_point(<\/span>aes(<\/span>color =<\/span> zvar))<\/span> +<\/span> scale_color_manual(<\/span>values =<\/span> c<\/span>(<\/span>\"orange\"<\/span>,<\/span> \n    \"purple\"<\/span>))<\/span> +<\/span> theme(<\/span>legend.position =<\/span> c<\/span>(<\/span>1<\/span>,<\/span> 1<\/span>),<\/span> legend.justification =<\/span> c<\/span>(<\/span>1<\/span>,<\/span> \n    1<\/span>))<\/span>\n\n# marginal density of x - plot on top<\/span>\nplot_top <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>xvar,<\/span> fill =<\/span> zvar))<\/span> +<\/span> geom_density(<\/span>alpha =<\/span> 0.5<\/span>)<\/span> +<\/span> \n    scale_fill_manual(<\/span>values =<\/span> c<\/span>(<\/span>\"orange\"<\/span>,<\/span> \"purple\"<\/span>))<\/span> +<\/span> theme(<\/span>legend.position =<\/span> \"none\"<\/span>)<\/span>\n\n# marginal density of y - plot on the right<\/span>\nplot_right <-<\/span> ggplot(<\/span>xy,<\/span> aes(<\/span>yvar,<\/span> fill =<\/span> zvar))<\/span> +<\/span> geom_density(<\/span>alpha =<\/span> 0.5<\/span>)<\/span> +<\/span> \n    coord_flip()<\/span> +<\/span> scale_fill_manual(<\/span>values =<\/span> c<\/span>(<\/span>\"orange\"<\/span>,<\/span> \"purple\"<\/span>))<\/span> +<\/span> theme(<\/span>legend.position =<\/span> \"none\"<\/span>)<\/span>\n\n# arrange the plots together, with appropriate height and width for each row<\/span>\n# and column<\/span>\ngrid.arrange(<\/span>plot_top,<\/span> empty,<\/span> scatter,<\/span> plot_right,<\/span> ncol =<\/span> 2<\/span>,<\/span> nrow =<\/span> 2<\/span>,<\/span> widths =<\/span> c<\/span>(<\/span>4<\/span>,<\/span> \n    1<\/span>),<\/span> heights =<\/span> c<\/span>(<\/span>1<\/span>,<\/span> 4<\/span>))<\/span>\n<\/code><\/pre>\n<\/div>\n
\"plot<\/p>\n
Original post is available here<\/a><\/p>\n<\/div>\n
\n
Home<\/a> Projects<\/a> ArtistD<\/a> CV<\/a> About<\/a><\/h4>\n
Felix Yanhui Fan, PhD \"Smiley<\/h3>\n
\"Smiley <\/a>\"Smiley <\/a>\"Smiley<\/a><\/p>\n
\n
Recent Posts<\/h3>\n