10.1 Histograms and density plots

Histograms and density plots are both visualizations of the distribution of quantitative variables.

10.1.1 Histogram

We will use a previous object to create a histogram, with geom_histogram. You can get it back with:

rnaseq2 <- read_csv("DataViz_source_files-main/files/GSE150029_rnaseq_log2_long.csv")

ggplot(rnaseq2, aes(x=log2_counts)) + 
  geom_histogram()

Split the histogram per sample:

ggplot(rnaseq2, aes(x=log2_counts, color=sample)) + 
  geom_histogram() 

Set position to “identity”: histograms will not be “on top of each other” but on a comparative scale:

ggplot(rnaseq2, aes(x=log2_counts, color=sample)) + 
  geom_histogram(position="identity")

Set alpha (transparency to 0.2):

ggplot(rnaseq2, aes(x=log2_counts, color=sample)) + 
  geom_histogram(position="identity", alpha=0.2)

Express as density instead of counts:

ggplot(rnaseq2, aes(x=log2_counts, color=sample)) + 
  geom_histogram(aes(y=after_stat(density)), position="identity", alpha=0.2) +
  theme_classic()

10.1.2 Density plot

It is similar to create a density plot, using geom_density:

ggplot(rnaseq2, aes(x=log2_counts, color=sample)) + 
  geom_density(alpha=0.2) +
  theme_classic()

You can use fill instead:

ggplot(rnaseq2, aes(x=log2_counts, fill=sample)) + 
  geom_density(alpha=0.2) +
  theme_classic()

10.1.3 Histogram + density

As we combined geom_boxplot and geom_violin in a previous exercise, we can also combine geom_histogram and geom_density:

ggplot(rnaseq2, aes(x=log2_counts, color=sample)) + 
  geom_histogram(aes(y=after_stat(density)), position="identity", alpha=0.2) +
  geom_density(alpha=0.2) +
  theme_classic()