How do I filter by multiple values of muliple categorical variables to make a plot in R?

Question

Suppose I have data like this:

type      source   weight
cabbage   store    2.2
cabbage   farm     2.3
cabbage   farm     1.9
celery    store    2.1
celery    farm     2.0
celery    store    1.7
turnip    farm     1.5
turnip    store    2.5

1) How do I make a boxplot of weights for cabbage and celery combined? Ie a single boxplot in which the data comes from the column weight , but only if the column type is "cabbage" or "celery".

2) How do I make a boxplot filtering by both categorical variables? Ie a single boxplot in which the data comes from the column weight , but only if the column type is "cabbage" or "celery" AND the column source is "farm".

Answer 1

Just provide boxplot with the filtered data as follows

df<-data.frame(type=c("cabbage","cabbage","cabbage","celery","celery","celery","turnip","turnip"), weight=c(2.2,2.3,1.9,2.1,2.0,1.7,1.5,2.5))
> df
     type weight
1 cabbage    2.2
2 cabbage    2.3
3 cabbage    1.9
4  celery    2.1
5  celery    2.0
6  celery    1.7
7  turnip    1.5
8  turnip    2.5
> boxplot(df$weight[df$type %in% c("cabbage","celery")])

This uses plain strings for type , but it will also work for factors.