Skip to content

Formula Grammar

This section of the documentation describes the formula grammar used by Formulaic. It is almost identical that used by patsy and R, and so most formulas should work without modification. However, there are some differences, which are called out below.

Operators

In this section, we introduce a complete list of the grammatical operators that you can use by default in your formulas. They are listed such that each section (demarcated by "-----") has higher precedence then the block that follows. When you write a formula involving several operators of different precedence, those with higher precedence will be resolved first. "Arity" is the number of arguments the operator takes. Within operators of the same precedence, all binary operators are evaluated from left to right (they are left-associative). To highlight differences in grammar betweeh formulaic, patsy and R, we highlight any differences below. If there is a checkmark the Formulaic, Patsy and R columns, then the grammar is consistent across all three, unless otherwise indicated.

Operator Arity Description Formulaic Patsy R
"..."1 1 String literal. ✓ ✓ ✗
[0-9]+\.[0-9]+1 1 Numerical literal. ✓ ✗ ✗
`...`1 1 Quotes fieldnames within the incoming dataframe, allowing the use of special characters, e.g. `my|special$column!` ✓ ✗ ✓
{...}1 1 Quotes python operations, as a more convenient way to do Python operations than I(...), e.g. {`my|col`**2} ✓ ✗ ✗
<function>(...)1 1 Python transform on column, e.g. my_func(x) which is equivalent to {my_func(x)} ✓2 ✓ ✗
-----
(...) 1 Groups operations, overriding normal precedence rules. All operations with the parentheses are performed before the result of these operations is permitted to be operated upon by its peers. ✓ ✓ ✓
-----
** 2 Includes all n-th order interactions of the terms in the left operand, where n is the (integral) value of the right operand, e.g. (a+b+c)**2 is equivalent to a + b + c + a:b + a:c + b:c. ✓ ✓ ✓
^ 2 Alias for **. ✓ ✗3 ✓
-----
: 2 Adds a new term that corresponds to the interaction of its operands (i.e. their elementwise product). ✓4 ✓ ✓
-----
* 2 Includes terms for each of the additive and interactive effects of the left and right operands, e.g. a * b is equivalent to a + b + a:b. ✓ ✓ ✓
/ 2 Adds terms describing nested effects. It expands to the addition of a new term for the left operand and the interaction of all left operand terms with the right operand, i.e a / b is equivalent to a + a:b, (a + b) / c is equivalent to a + b + a:b:c, and a/(b+c) is equivalent to a + a:b + a:c.5 ✓ ✓ ✓
%in% 2 As above, but with arguments inverted: e.g. b %in% a is equivalent to a / b. ✓ ✗ ✓
-----
+ 2 Adds a new term to the set of features. ✓ ✓ ✓
- 2 Removes a term from the set of features (if present). ✓ ✓ ✓
+ 1 Returns the current term unmodified (not very useful). ✓ ✓ ✓
- 1 Negates a term (only implemented for 0, in which case it is replaced with 1). ✓ ✓ ✓
-----
| 2 Splits a formula into multiple parts, allowing the simultaneous generation of multiple model matrices. When on the right-hand-side of the ~ operator, all parts will attract an additional intercept term by default. ✓ ✗ ✓6
-----
~ 1,2 Separates the target features from the input features. If absent, it is assumed that we are considering only the the input features. Unless otherwise indicated, it is assumed that the input features implicitly include an intercept. ✓ ✓ ✓

Transforms

Formulaic supports arbitrary transforms, any of which can also preserve state so that new data can undergo the same transformation as that used during modelling. The currently implemented transforms are shown below. Commonly used transforms that have not been implemented by formulaic are explicitly noted also.

Transform Description Formulaic Patsy R
I(...) Identity transform, allowing arbitrary Python/R operations, e.g. I(x+y). Note that in formulaic, it is more idiomatic to use {x+y}. ✓ ✓ ✓
Q('<column_name>') Look up feature by potentially exotic name, e.g. Q('wacky name!'). Note that in formulaic, it is more idiomatic to use `wacky name!`. ✓ ✓ ✗
C(...) Categorically encode a column, e.g. C(x) ✓ ✓ ✓
center(...) Shift column data so mean is zero. ✓ ✓ ✗
scale(...) Shift column so mean is zero and variance is 1. ✓ ✓7 ✓
standardize(...) Alias of scale. ✓8 ✓ ✗
poly(...) Generates a polynomial basis, allowing non-linear fits. ✓ ✗ ✓
bs(...) Generates a B-Spline basis, allowing non-linear fits. ✓ ✓ ✓
cr(...) Generates a natural cubic spline basis, allowing non-linear fits. ✗ ✓ ✓
cc(...) Generates a cyclic cubic spline basis, allowing non-linear fits. ✗ ✓ ✓
te(...) Generates a tensor product smooth. ✗ ✓ ✓
hashed(...) Categorically encode a deterministic hash of a column. ✓ ✗ ✗
... Others? Contributions welcome! ? ? ?

Tip

Any function available in the context dictionary will also be available as transform, along with some commonly used functions imported from numpy: log, log10, log2, exp, exp10, and exp2. In addition the numpy module is always available as np. Thus, formulas like: log(y) ~ x + 10 will always do the right thing, even when these functions have not been made available in the user namespace.

Note

Formulaic does not (yet) support including extra terms in the formula that will not result in additions to the dataframe, for example model annotations like R's offset(...).

Behaviours and Conventions

Beyond the formula operator grammar itself there are some differing behaviours and conventions of which you should be aware.

  • Formulaic follows Patsy and then enhanced Formula R package in that both sides of the ~ operator are treated considered to be using the formula grammar, with the only difference being that the right hand side attracts an intercept by default. In vanilla R, the left hand side is treated as R code (and so x + y ~ z would result in a single column on the left-hand-side). You can recover vanilla R's behaviour by nesting the operations in a Python operator block (as described in the operator table): {y1 + y2} ~ a + b.
  • Formula terms in Formulaic are always sorted first by the order of the interaction, and then alphabetically. In R and patsy, this second ordering is done in the order that columns were introduced to the formula (patsy additionally sorts by which fields are involved in the interactions). As a result formulas generated by formulaic with the same set of fields will always generate the same model matrix.
  • Formulaic follows patsy's more rigourous handling of whether or not to include an intercept term. In R, b-1 and (b-1) both do not have an intercept, whereas in Formulaic and Patsy the parentheses are resolved first, and so the first does not have an intercept and the second does (because '1 +' is implicitly prepended to the right hand side of the formula).
  • Formulaic borrows a clever algorithm introduced by Patsy to carefully choose where to reduce the rank of the model matrix in order to ensure that the matrix is structurally full rank. This avoids producing over-specified model matrices in contexts that R would (since it only considers local full-rank structure, rather than global structure). You can read more about this in Patsy's documentation.

  1. This "operator" is actually part of the tokenisation process. ↩↩↩↩↩

  2. Formulaic additionally supports quoted fields with special characters, e.g. my_func(`my|special+column`). ↩

  3. The caret operator is not supported, but will not cause an error. It is ignored by the patsy formula parser, and treated as XOR Python operation on column. ↩

  4. Note that Formulaic also allows you to use this to scale columns, for example: 2.5:a (this scaling happens after factor coding). ↩

  5. This somewhat confusing operator is useful when you want to include hierachical features in your data, and where certain interaction terms do not make sense (particularly in ANOVA contexts). For example, if a represents countries, and b represents cities, then the full product of terms from a * b === a + b + a:b does not make sense, because any value of b is guaranteed to coincide with a value in a, and does not independently add value. Thus, the operation a / b === a + a:b results in more sensible dataset. As a result, the / operator is right-distributive, since if b and c were both nested in a, you would want a/(b+c) === a + a:b + a:c. Likewise, the operator is not left-distributive, since if c is nested under both a and b separately, then you want (a + b)/c === a + b + a:b:c. Lastly, if c is nested in b, and b is nested in a, then you would want a/b/c === a + a:(b/c) === a + a:b + a:b:c. ↩

  6. Implemented by an R package called Formula that extends the default formula syntax. ↩

  7. Patsy uses the rescale keyword rather than scale, but provides the same functionality. ↩

  8. For increased compatibility with patsy, we use patsy's signature for standardize. ↩