Formula

  1. Create
  2. Test
  3. Get
  4. Update
  5. Evaluate
  6. Convert

Create

What How Details
With intercept and single variable y ~ x  
Without intercept, and single variable y ~ -1 + x  
With all variables y ~ .  
From string as.formula(str)  
From string eval(parse(text='y ~ x')) Not recommended

Test

What How Details
Is formula is(f, 'formula')  
Formulas are equal f1 == f2  
Formulas are equal isTRUE(all.equal(f1, f2))  

Get

What How Details
Variable names all.vars(f)  
Left-hand side variable names all.vars(update(y ~ x, ~ 1))  
Right-hand side variable names all.vars(update(y ~ x, 1 ~ .))  
Environment environment(f)  

Update

What How Details
Change terms update(y~x, ~ w + z)  
Append terms update(y~x, ~ . + w)  
Change response update(y~x, z ~ .)  
Drop response formula(delete.response(terms(y ~ x)))  
Remove the environment environment(f) = NULL  

Evaluate

What How Details
Evaluate a formula with known coefficients mm = model.matrix( ~ a + poly(b, 2, raw=TRUE), data.frame(a=1, b=2:4))
pred = coef %*% t(mm)		  

Convert

What How Details
String deparse(f)  
String array as.character(y~x)