Automatic handling of missing values

[christophscheuch]

For those stats which require complete data, missing values will be automatically removed with a warning in ´ggplot2´ (see here for the corresponding convenience function). The Python library plotnine also removes missing values with a warning. When I started creating plots with TidierPlots, I was confused by the axis labels until I realized that it automatically adds a missing category to the labels.

Here is a specific example:

using Tidier, PalmerPenguins

penguins = DataFrame(PalmerPenguins.load())

# Creating a plot without dropping missing variables leads to messed up axis labels
@ggplot(penguins, aes(x = bill_length_mm, y = bill_depth_mm)) +
  @geom_point()

# Dropping missing values from aes variables leads to proper labels
penguins_sub = @drop_na(penguins, bill_length_mm, bill_depth_mm)
@ggplot(penguins_sub, aes(x = bill_length_mm, y = bill_depth_mm)) +
  @geom_point()

Is this desired behavior or do we also want to have automatic handling of missing values eventually?

[rdboyes]

fixed in 2c45592