def test_variable_unrecognized_type():
with pytest.raises(ValueError):
Variable("x").set_type({"x": 1})
with pytest.raises(ValueError):
Variable("x").set_type({"x": [1, 2]})
with pytest.raises(ValueError):
Variable("x").set_type({"x": {"a": 1}})
def test_subset_index():
desc = model_description("threecats['b'] ~ continuous + dummy")
comp = Model(
Intercept(),
Term(Variable("continuous")),
Term(Variable("dummy")),
)
comp.add_response(Response(Term(Variable("threecats", level="b"))))
assert desc == comp
def test_term():
desc = model_description("x")
comp = Model(Intercept(), Term(Variable("x")))
assert desc == comp
desc = model_description("term_name_abc")
comp = Model(Intercept(), Term(Variable("term_name_abc")))
assert desc == comp
desc = model_description("`$%!N4m3##!! NNN`")
comp = Model(Intercept(), Term(Variable("$%!N4m3##!! NNN")))
assert desc == comp
def test_variable_set_data_errors():
x = Variable("x")
with pytest.raises(ValueError):
x.set_data(True)
with pytest.raises(ValueError):
x.kind = "hello"
x.set_data(True)
with pytest.raises(Exception):
x.kind = "categoric"
x.set_data(True)
def visitLiteralExpr(self, expr):
if expr.value == 0:
return NegatedIntercept()
elif expr.value == 1:
return Intercept()
else:
return Term(Variable(expr.value))
def test_term_new_data_numeric():
data = pd.DataFrame({"x": [10, 10, 10]})
var_expr = Parser(Scanner("x").scan(False)).parse()
var_term = Variable(var_expr.name.lexeme, var_expr.level)
var_term.set_type(data)
var_term.set_data()
assert (var_term.value == [10, 10, 10]).all()
data = pd.DataFrame({"x": [1, 2, 3]})
assert (var_term.eval_new_data(data).T == [1, 2, 3]).all()
def test_variable_eval_numeric():
x = Variable("x")
arr = np.array([1, 2, 3, 4])
series = pd.Series([1, 2, 3, 4])
x.eval_numeric(arr)
assert np.array_equal(x.value, arr)
x.eval_numeric(series)
assert np.array_equal(x.value, arr)
with pytest.raises(ValueError):
x.eval_numeric([1, 2, 3])
def test_variable_str():
assert str(Variable("a")) == "Variable(a)"
assert str(Variable("a", "hi")) == "Variable(a, reference='hi')"
assert repr(Variable("a")) == str(Variable("a"))
assert repr(Variable("a", "hi")) == str(Variable("a", "hi"))
def test_term_remove():
desc = model_description("x - y")
comp = Model(Intercept(), Term(Variable("x")))
assert desc == comp
desc = model_description("x - 5")
comp = Model(Intercept(), Term(Variable("x")))
assert desc == comp
desc = model_description("x - f(x)")
comp = Model(Intercept(), Term(Variable("x")))
assert desc == comp
desc = model_description("x - y:z")
comp = Model(Intercept(), Term(Variable("x")))
assert desc == comp
desc = model_description("x - (1|g)")
comp = Model(Intercept(), Term(Variable("x")))
assert desc == comp
desc = model_description("x - (z + y)")
comp = Model(Intercept(), Term(Variable("x")))
assert desc == comp
def test_term_interaction():
desc = model_description("x:y")
comp = Model(Intercept(), Term(Variable("x"), Variable("y")))
assert desc == comp
with pytest.raises(TypeError):
model_description("x:5")
desc = model_description("x:f(x)")
comp = Model(
Intercept(),
Term(Variable("x"), Call(LazyCall("f", [LazyVariable("x")], {}))),
)
assert desc == comp
desc = model_description("x:y:z")
comp = Model(Intercept(), Term(Variable("x"), Variable("y"),
Variable("z")))
assert desc == comp
desc = model_description("x:y*z")
comp = Model(
Intercept(),
Term(Variable("x"), Variable("y")),
Term(Variable("z")),
Term(Variable("x"), Variable("y"), Variable("z")),
)
assert desc == comp
# Note the parenthesis, here `*` resolves earlier than `:`
desc = model_description("x:(y*z)")
comp = Model(
Intercept(),
Term(Variable("x"), Variable("y")),
Term(Variable("x"), Variable("z")),
Term(Variable("x"), Variable("y"), Variable("z")),
)
assert desc == comp
with pytest.raises(TypeError):
model_description("x:(1|g)")
desc = model_description("x:(z + y)")
comp = Model(
Intercept(),
Term(Variable("x"), Variable("z")),
Term(Variable("x"), Variable("y")),
)
assert desc == comp
def test_term_add():
desc = model_description("x + y")
comp = Model(Intercept(), Term(Variable("x")), Term(Variable("y")))
assert desc == comp
desc = model_description("x + 5")
comp = Model(Intercept(), Term(Variable("x")), Term(Variable(5)))
assert desc == comp
desc = model_description("x + f(x)")
comp = Model(
Intercept(),
Term(Variable("x")),
Term(Call(LazyCall("f", [LazyVariable("x")], {}))),
)
assert desc == comp
desc = model_description("x + y:z")
comp = Model(Intercept(), Term(Variable("x")),
Term(Variable("y"), Variable("z")))
assert desc == comp
desc = model_description("x + (1|g)")
comp = Model(Intercept(), Term(Variable("x")),
GroupSpecificTerm(Intercept(), Term(Variable("g"))))
assert desc == comp
desc = model_description("x + (z + y)")
comp = Model(Intercept(), Term(Variable("x")), Term(Variable("z")),
Term(Variable("y")))
assert desc == comp
def test_group_specific_interactions():
desc = model_description("0 + (a*b|h+g)")
comp = Model(
GroupSpecificTerm(expr=Intercept(), factor=Term(Variable("h"))),
GroupSpecificTerm(expr=Intercept(), factor=Term(Variable("g"))),
GroupSpecificTerm(expr=Term(Variable("a")),
factor=Term(Variable("h"))),
GroupSpecificTerm(expr=Term(Variable("a")),
factor=Term(Variable("g"))),
GroupSpecificTerm(expr=Term(Variable("b")),
factor=Term(Variable("h"))),
GroupSpecificTerm(expr=Term(Variable("b")),
factor=Term(Variable("g"))),
GroupSpecificTerm(expr=Term(Variable("a"), Variable("b")),
factor=Term(Variable("h"))),
GroupSpecificTerm(Term(Variable("a"), Variable("b")),
factor=Term(Variable("g"))),
)
assert desc == comp
desc = model_description("0 + (0 + a*b|h+g)")
comp = Model(
GroupSpecificTerm(expr=Term(Variable("a")),
factor=Term(Variable("h"))),
GroupSpecificTerm(expr=Term(Variable("a")),
factor=Term(Variable("g"))),
GroupSpecificTerm(expr=Term(Variable("b")),
factor=Term(Variable("h"))),
GroupSpecificTerm(expr=Term(Variable("b")),
factor=Term(Variable("g"))),
GroupSpecificTerm(expr=Term(Variable("a"), Variable("b")),
factor=Term(Variable("h"))),
GroupSpecificTerm(Term(Variable("a"), Variable("b")),
factor=Term(Variable("g"))),
)
assert desc == comp
def test_term_slash():
desc = model_description("x / y")
comp = Model(Intercept(), Term(Variable("x")),
Term(Variable("x"), Variable("y")))
assert desc == comp
with pytest.raises(TypeError):
model_description("x / 5")
desc = model_description("x / f(x)")
comp = Model(
Intercept(),
Term(Variable("x")),
Term(Variable("x"), Call(LazyCall("f", [LazyVariable("x")], {}))),
)
assert desc == comp
desc = model_description("x / y:z")
comp = Model(
Intercept(),
Term(Variable("x")),
Term(Variable("x"), Variable("y"), Variable("z")),
)
assert desc == comp
with pytest.raises(TypeError):
model_description("x / (1|g)")
desc = model_description("x / (z + y)")
comp = Model(
Intercept(),
Term(Variable("x")),
Term(Variable("x"), Variable("z")),
Term(Variable("x"), Variable("y")),
)
assert desc == comp
def test_term_new_data_categoric():
data = pd.DataFrame({"x": ["A", "B", "C"]})
# Full rank encoding
var_expr = Parser(Scanner("x").scan(False)).parse()
var_term = Variable(var_expr.name.lexeme, var_expr.level)
var_term.set_type(data)
var_term.set_data(spans_intercept=True)
assert (np.array([[1, 0, 0], [0, 1, 0], [0, 0, 1]]) == var_term.value).all()
data = pd.DataFrame({"x": ["B", "C"]})
assert (var_term.eval_new_data(data) == np.array([[0, 1, 0], [0, 0, 1]])).all()
# It remembers it saw "A", "B", and "C", but not "D".
# So when you pass a new level, it raises a ValueError.
with pytest.raises(
ValueError, match="The levels D in 'x' are not present in the original data set."
):
data = pd.DataFrame({"x": ["B", "C", "D"]})
var_term.eval_new_data(data)
# The same with reduced encoding
data = pd.DataFrame({"x": ["A", "B", "C"]})
var_expr = Parser(Scanner("x").scan(False)).parse()
var_term = Variable(var_expr.name.lexeme, var_expr.level)
var_term.set_type(data)
var_term.set_data()
assert (np.array([[0, 0], [1, 0], [0, 1]]) == var_term.value).all()
data = pd.DataFrame({"x": ["A", "C"]})
assert (var_term.eval_new_data(data) == np.array([[0, 0], [0, 1]])).all()
# It remembers it saw "A", "B", and "C", but not "D".
# So when you pass a new level, it raises a ValueError.
with pytest.raises(
ValueError, match="The levels D in 'x' are not present in the original data set."
):
data = pd.DataFrame({"x": ["B", "C", "D"]})
var_term.eval_new_data(data)
def visitQuotedNameExpr(self, expr):
# Quoted names don't accept levels yet.
return Term(Variable(expr.expression.lexeme[1:-1]))
def visitVariableExpr(self, expr):
if expr.level:
level = expr.level.value
else:
level = None
return Term(Variable(expr.name.lexeme, level))
