Take a look: How to remove NA (data with missing values) in geom_col? - #2 by Flm
The easiest way is to change the first line ggplot(df_plot, aes(...)) into
df_plot %>%
drop_na() %>%
ggplot(aes(x = df_plot$hour, y = df_plot$`amount of samples`, fill = df_plot$system)) +