Geoms and Aesthetics

 

These exercises cover the Geoms and Aesthetics of ggplot2 for Plotting in R.

These first few exercises will run through some of the simple principles of creating a ggplot2 object, assigning aesthetics mappings and geoms.

Exercise 1 - Scatterplots

• Read in the cleaned patients dataset as we saw in ggplot2 course earlier (“patients_clean_ggplot2.txt”)

• Using the patient data set generate a scatter plot of BMI versus Weight.

• Extend the plot by adding a color scale to the scatter plot based on the Height variable.

• Using an additional geom, add an extra layer of a fit line to our scatter plot.

## `geom_smooth()` using method = 'loess' and formula = 'y ~ x'

## Warning: The following aesthetics were dropped during statistical transformation:
## colour.
## i This can happen when ggplot fails to infer the correct grouping structure in
##   the data.
## i Did you forget to specify a `group` aesthetic or to convert a numerical
##   variable into a factor?

• Does the fit in look good? Look at the description for ?geom_smooth() and adjust the method for a better fit.

## `geom_smooth()` using formula = 'y ~ x'

## Warning: The following aesthetics were dropped during statistical transformation:
## colour.
## i This can happen when ggplot fails to infer the correct grouping structure in
##   the data.
## i Did you forget to specify a `group` aesthetic or to convert a numerical
##   variable into a factor?

Exercise 2 - Boxplots, Violin and Jitter plots

• Generate a boxplot of BMIs comparing smokers and non-smokers.

• Following from the boxplot comparing smokers and non-smokers, color the boxplot edges by Sex.

• Create a violin plot comparing the BMI between people with different pets.

## Warning: Groups with fewer than two datapoints have been dropped.
## i Set `drop = FALSE` to consider such groups for position adjustment purposes.
## Groups with fewer than two datapoints have been dropped.
## i Set `drop = FALSE` to consider such groups for position adjustment purposes.

• Try out a different smoothing kernel for estimating the density. Use ?geom_violin find out more.

## Warning: Groups with fewer than two datapoints have been dropped.
## i Set `drop = FALSE` to consider such groups for position adjustment purposes.
## Groups with fewer than two datapoints have been dropped.
## i Set `drop = FALSE` to consider such groups for position adjustment purposes.

• Add both different color fills for each violin, and a jitter plot overlay.

## Warning: Groups with fewer than two datapoints have been dropped.
## i Set `drop = FALSE` to consider such groups for position adjustment purposes.
## Groups with fewer than two datapoints have been dropped.
## i Set `drop = FALSE` to consider such groups for position adjustment purposes.

• Reduce the amount of jitter, so that points are a little more narrow.

## Warning: Groups with fewer than two datapoints have been dropped.
## i Set `drop = FALSE` to consider such groups for position adjustment purposes.
## Groups with fewer than two datapoints have been dropped.
## i Set `drop = FALSE` to consider such groups for position adjustment purposes.

Exercise 3 - Histogram and Density plots

• Generate a histogram of BMIs with each bar colored blue.

## `stat_bin()` using `bins = 30`. Pick better value with `binwidth`.

• Generate density plots of BMIs colored by Sex. Make sure you can see both distributions by adjusting the transparency.

HINT: alpha can be used to control transparency.

Exercise 4 - GGplot object

• Take your last plot object and extract out the data from it.

##            ID      Name   Race Age    Sex     Smokes   Height   Weight
## 1 Patient-001 Demetrius  White  44   Male     Smoker 176.4791 77.07070
## 2 Patient-002   Rosario  White  43   Male     Smoker 174.7485 83.26835
## 3 Patient-003     Julio  Black  44   Male Non-Smoker 173.9449 85.63231
## 4 Patient-004      Lupe  White  44   Male Non-Smoker 180.0289 88.72032
## 5 Patient-005    Lavern    Cat  43   Male Non-Smoker 178.6365 97.05405
## 6 Patient-006    Bernie Native  43 Female Non-Smoker 159.7778 68.31247
##        Birth         State Pet Grade_Level  Died      Count       Date Grade
## 1 1972-02-11      New York Cat           1  TRUE -0.7440011 2015-03-04     1
## 2 1972-07-22       Florida Cat           1  TRUE -1.0714531 2015-03-04     1
## 3 1971-11-23   Connecticut Dog           2 FALSE -0.8215636 2015-03-04     2
## 4 1971-10-03 Massachusetts Dog           1  TRUE  0.4976671 2015-03-04     1
## 5 1972-11-23        Kansas Cat           3 FALSE -1.2099217 2015-03-04     3
## 6 1972-07-30      Illinois Cat           2 FALSE  0.5241618 2015-03-04     2
##        BMI Overweight
## 1 24.74586      FALSE
## 2 27.26799       TRUE
## 3 28.30182       TRUE
## 4 27.37403       TRUE
## 5 30.41397       TRUE
## 6 26.75882       TRUE

• Now check the aesthetic mappings

## Aesthetic mapping: 
## * `x` -> `BMI`