Templates for graphics.

This document, which will grow as we encounter new graphical patterns, contains templates for ggplot() graphics.

To use the template, create your glyph-ready data, and put in your own variables in place of the ones listed here IN CAPS.

Scatter Plots

Template

Make a scatterplot with some bells and whistles.

YOUR_DATA_TABLE %>%              # REQUIRED
  ggplot(aes(x=X_VAR,y=Y_VAR)) + # REQUIRED -- insert your variable names
  geom_point() +                 # REQUIRED
  aes(colour=GROUP_VAR) +        # Optional -- map variable onto color
  aes(size=SIZE_VAR) +           # Optional -- map variable onto size
  scale_x_log10() +              # Optional -- use logarithmic scale for x axis.  
  scale_y_log10() +              # Optional -- use logarithmic y scale for y-axis
  facet_wrap(~FACET_VAR, ncol=4) # Optional -- faceting variable

You can delete any of the optional lines if you don’t want that aspect of the plot. Then, replace each of the placeholders — YOUR_DATA_TABLE, X_VAR, Y_VAR, GROUP_VAR, FACET_VAR — with the relevant variable from your data table. (If you delete an optional line, you don’t need to worry about the placeholder variable; it will disappear from the statement.)

Example

# Create the glyph-ready data
data( NHANES )
Small <- sample_n( NHANES, size=2000 ) # glyph ready table

# Template pasted in and placeholders filled in with real variables.
Small %>%                        # REQUIRED
  ggplot(aes(x=age,y=height)) +  # REQUIRED -- insert your variable names
  geom_point() +                 # REQUIRED
  aes(colour=smoker) +           # Optional -- map variable onto color
  aes(size=diabetic) +           # Optional -- map variable onto size
  scale_x_log10() +              # Optional -- use logarithmic scale for x axis.  
  facet_wrap(~sex, ncol=4)       # Optional -- faceting variable

One of the optional lines (scale_y_log10()) was deleted so the y-axis is on the usual linear scale.

Bar Charts

[Under construction … I need a better example]

Template

DATA_TABLE %>%                           # REQUIRED
  ggplot(aes(x=X_VAR,                    # REQUIRED
             y=Y_VAL, 
             fill=FILL_VAR)) +           # REQUIRED
  geom_bar(stat='identity',              # REQUIRED
           position=                     # REQUIRED
             position_HOW(width=.9)) +   # REQUIRED options: stack, proportion, dodge
  facet_wrap(~FACET_VAR,ncol=3)  +       # Optional

Note the HOW placeholder. Choose one of stack, proportion, or dodge so that the function looks like position_stack(), position_proportion(), or position_dodge().

Example

Make your glyph-ready data …

FirstPlaceTally <-
  Minneapolis2013 %>%
  group_by( First ) %>%
  summarise( total=n() )
ByPrecinct <-
  FirstPlaceTally %>% 
  filter(total > 5000) %>% # Just the big candidates.
  inner_join( Minneapolis2013, . ) %>%
  group_by( Precinct, First) %>% 
  summarise( total=n() ) %>%
  mutate( First=droplevels(First) )

Copy and customize the template …

ByPrecinct %>% 
  ggplot(aes(x=Precinct,
             y=total,
             fill=First)) +
  geom_bar(stat='identity', 
           position=position_stack(width=.9)) + 
  facet_wrap(~First,ncol=3) 

---

Please use the comment system to make suggestions, point out errors, or to discuss the topic.

comments powered by Disqus

Written by Daniel Kaplan for the Data & Computing Fundamentals Course. Development was supported by grants from the National Science Foundation for Project Mosaic (NSF DUE-0920350) and from the Howard Hughes Medical Institute.