Skip to content

Test with PyMC v6 #33

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
michaelosthege merged 1 commit into from
May 23, 2026
Merged

Test with PyMC v6 #33

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion .github/workflows/pipeline.yml
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -11,7 +11,7 @@ jobs:
strategy:
matrix:
python-version: ["3.11", "3.14"]
pymc-version: ["without", "'pymc>=5.0.0'"]
pymc-version: ["without", "'pymc>=6.0.0'"]
defaults:
run:
shell: bash -l {0}
Expand Down
212 changes: 104 additions & 108 deletions murefi/test_core.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -799,10 +799,9 @@ class TestSymbolicComputation:
@pytest.mark.skipif(not HAS_PYMC, reason="requires PyMC")
def test_timeseries_support(self):
t = numpy.linspace(0, 10, 10)
with _backend.config.change_flags(compute_test_value="off"):
y = pt.scalar("TestY", dtype=_backend.config.floatX)
assert isinstance(y, pt.TensorVariable)
ts = murefi.Timeseries(t, y, independent_key="Test", dependent_key="Test")
y = pt.scalar("TestY", dtype=_backend.config.floatX)
assert isinstance(y, pt.TensorVariable)
ts = murefi.Timeseries(t, y, independent_key="Test", dependent_key="Test")
return

@pytest.mark.skipif(not HAS_PYMC, reason="requires PyMC")
Expand All @@ -814,131 +813,128 @@ def test_symbolic_parameter_mapping(self, df_mapping):
)

# create a theta that is a mix of constant and symbolic variables
with _backend.config.change_flags(compute_test_value="off"):
theta_fitted = [
1,
pt.scalar("mu_max", dtype=_backend.config.floatX),
13,
pt.scalar("t_acc", dtype=_backend.config.floatX),
]

# map it to the two replicates
expected = {
"A01": numpy.array([1.0, None, 3.0, 4.0, 5.0, 6.0, 7.0]),
"B02": numpy.array([11.0, None, 13.0, None, 15.0, 16.0, 17.0]),
}
mapped = parmap.repmap(theta_fitted)
assert mapped.keys() == expected.keys()
for rid in expected.keys():
for exp, act in zip(expected[rid], mapped[rid]):
if exp is not None:
assert exp == act
else:
assert isinstance(act, pt.TensorVariable)
theta_fitted = [
1,
pt.scalar("mu_max", dtype=_backend.config.floatX),
13,
pt.scalar("t_acc", dtype=_backend.config.floatX),
]

# map it to the two replicates
expected = {
"A01": numpy.array([1.0, None, 3.0, 4.0, 5.0, 6.0, 7.0]),
"B02": numpy.array([11.0, None, 13.0, None, 15.0, 16.0, 17.0]),
}
mapped = parmap.repmap(theta_fitted)
assert mapped.keys() == expected.keys()
for rid in expected.keys():
for exp, act in zip(expected[rid], mapped[rid]):
if exp is not None:
assert exp == act
else:
assert isinstance(act, pt.TensorVariable)
return

@pytest.mark.skipif(not HAS_PYMC, reason="requires PyMC")
def test_symbolic_predict_replicate(self):
with _backend.config.change_flags(compute_test_value="off"):
inputs = [
pt.scalar("beta", dtype=_backend.config.floatX),
pt.scalar("A", dtype=_backend.config.floatX),
]
ode_parameters = [0.23, inputs[0]]
y0 = [inputs[1], 2.0, 0.0]
t = numpy.linspace(0, 1, 5)
model = _mini_model()
inputs = [
pt.scalar("beta", dtype=_backend.config.floatX),
pt.scalar("A", dtype=_backend.config.floatX),
]
ode_parameters = [0.23, inputs[0]]
y0 = [inputs[1], 2.0, 0.0]
t = numpy.linspace(0, 1, 5)
model = _mini_model()

template = murefi.Replicate("TestRep")
# one observation of A, two observations of C
template["A"] = murefi.Timeseries(t[:3], [0] * 3, independent_key="A", dependent_key="A")
template["C1"] = murefi.Timeseries(t[2:4], [0] * 2, independent_key="C", dependent_key="C1")
template["C2"] = murefi.Timeseries(t[1:4], [0] * 3, independent_key="C", dependent_key="C2")
template = murefi.Replicate("TestRep")
# one observation of A, two observations of C
template["A"] = murefi.Timeseries(t[:3], [0] * 3, independent_key="A", dependent_key="A")
template["C1"] = murefi.Timeseries(t[2:4], [0] * 2, independent_key="C", dependent_key="C1")
template["C2"] = murefi.Timeseries(t[1:4], [0] * 3, independent_key="C", dependent_key="C2")

# construct the symbolic computation graph
prediction = model.predict_replicate(y0 + ode_parameters, template)
# construct the symbolic computation graph
prediction = model.predict_replicate(y0 + ode_parameters, template)

assert isinstance(prediction, murefi.Replicate)
assert prediction.rid == "TestRep"
assert "A" in prediction
assert "B" not in prediction
assert "C1" in prediction
assert "C2" in prediction
assert isinstance(prediction, murefi.Replicate)
assert prediction.rid == "TestRep"
assert "A" in prediction
assert "B" not in prediction
assert "C1" in prediction
assert "C2" in prediction

assert isinstance(prediction["A"].y, pt.TensorVariable)
assert isinstance(prediction["C1"].y, pt.TensorVariable)
assert isinstance(prediction["C2"].y, pt.TensorVariable)
assert isinstance(prediction["A"].y, pt.TensorVariable)
assert isinstance(prediction["C1"].y, pt.TensorVariable)
assert isinstance(prediction["C2"].y, pt.TensorVariable)

outputs = [prediction["A"].y, prediction["C1"].y, prediction["C2"].y]
outputs = [prediction["A"].y, prediction["C1"].y, prediction["C2"].y]

# compile a PyTensor function for performing the computation
f = _backend.function(inputs, outputs)
# compile a PyTensor function for performing the computation
f = _backend.function(inputs, outputs)

# compute the model outcome
actual = f(0.85, 2.0)
# compute the model outcome
actual = f(0.85, 2.0)

assert numpy.allclose(actual[0], [2.0, 1.4819299, 1.28322046])
assert numpy.allclose(actual[1], [0.71677954, 0.83004323])
assert numpy.allclose(actual[2], [0.5180701, 0.71677954, 0.83004323])
assert numpy.allclose(actual[0], [2.0, 1.4819299, 1.28322046])
assert numpy.allclose(actual[1], [0.71677954, 0.83004323])
assert numpy.allclose(actual[2], [0.5180701, 0.71677954, 0.83004323])
return

@pytest.mark.skipif(not HAS_PYMC, reason="requires PyMC")
def test_symbolic_predict_dataset(self):
with _backend.config.change_flags(compute_test_value="off"):
inputs = [
pt.scalar("beta", dtype=_backend.config.floatX),
pt.scalar("A", dtype=_backend.config.floatX),
]
ode_parameters = [0.23, inputs[0]]
y0 = [inputs[1], 2.0, 0.0]
t = numpy.linspace(0, 1, 5)
model = _mini_model()

# create a parameter mapping
mapping = pandas.DataFrame(columns="rid,A0,B0,C0,alpha,beta".split(",")).set_index("rid")
mapping.loc["TestRep"] = "A0,B0,C0,alpha,beta".split(",")
pmap = murefi.ParameterMapping(mapping, bounds=dict(), guesses=dict())
assert pmap.ndim == 5
numpy.testing.assert_array_equal(tuple(pmap.parameters.keys()), "A0,B0,C0,alpha,beta".split(","))

# create a dataset
ds_template = murefi.Dataset()

# One replicate with one observation of A, two observations of C
template = murefi.Replicate("TestRep")
template["A"] = murefi.Timeseries(t[:3], [0] * 3, independent_key="A", dependent_key="A")
template["C1"] = murefi.Timeseries(t[2:4], [0] * 2, independent_key="C", dependent_key="C1")
template["C2"] = murefi.Timeseries(t[1:4], [0] * 3, independent_key="C", dependent_key="C2")
ds_template["TestRep"] = template

# construct the symbolic computation graph
prediction = model.predict_dataset(ds_template, pmap, parameters=y0 + ode_parameters)

assert isinstance(prediction, murefi.Dataset)
assert "A" in prediction["TestRep"]
assert "B" not in prediction["TestRep"]
assert "C1" in prediction["TestRep"]
assert "C2" in prediction["TestRep"]
inputs = [
pt.scalar("beta", dtype=_backend.config.floatX),
pt.scalar("A", dtype=_backend.config.floatX),
]
ode_parameters = [0.23, inputs[0]]
y0 = [inputs[1], 2.0, 0.0]
t = numpy.linspace(0, 1, 5)
model = _mini_model()

assert isinstance(prediction["TestRep"]["A"].y, pt.TensorVariable)
assert isinstance(prediction["TestRep"]["C1"].y, pt.TensorVariable)
assert isinstance(prediction["TestRep"]["C2"].y, pt.TensorVariable)
# create a parameter mapping
mapping = pandas.DataFrame(columns="rid,A0,B0,C0,alpha,beta".split(",")).set_index("rid")
mapping.loc["TestRep"] = "A0,B0,C0,alpha,beta".split(",")
pmap = murefi.ParameterMapping(mapping, bounds=dict(), guesses=dict())
assert pmap.ndim == 5
numpy.testing.assert_array_equal(tuple(pmap.parameters.keys()), "A0,B0,C0,alpha,beta".split(","))

outputs = [
prediction["TestRep"]["A"].y,
prediction["TestRep"]["C1"].y,
prediction["TestRep"]["C2"].y,
]
# create a dataset
ds_template = murefi.Dataset()

# compile a PyTensor function for performing the computation
f = _backend.function(inputs, outputs)
# One replicate with one observation of A, two observations of C
template = murefi.Replicate("TestRep")
template["A"] = murefi.Timeseries(t[:3], [0] * 3, independent_key="A", dependent_key="A")
template["C1"] = murefi.Timeseries(t[2:4], [0] * 2, independent_key="C", dependent_key="C1")
template["C2"] = murefi.Timeseries(t[1:4], [0] * 3, independent_key="C", dependent_key="C2")
ds_template["TestRep"] = template

# compute the model outcome
actual = f(0.85, 2.0)
# construct the symbolic computation graph
prediction = model.predict_dataset(ds_template, pmap, parameters=y0 + ode_parameters)

assert numpy.allclose(actual[0], [2.0, 1.4819299, 1.28322046])
assert numpy.allclose(actual[1], [0.71677954, 0.83004323])
assert numpy.allclose(actual[2], [0.5180701, 0.71677954, 0.83004323])
assert isinstance(prediction, murefi.Dataset)
assert "A" in prediction["TestRep"]
assert "B" not in prediction["TestRep"]
assert "C1" in prediction["TestRep"]
assert "C2" in prediction["TestRep"]

assert isinstance(prediction["TestRep"]["A"].y, pt.TensorVariable)
assert isinstance(prediction["TestRep"]["C1"].y, pt.TensorVariable)
assert isinstance(prediction["TestRep"]["C2"].y, pt.TensorVariable)

outputs = [
prediction["TestRep"]["A"].y,
prediction["TestRep"]["C1"].y,
prediction["TestRep"]["C2"].y,
]

# compile a PyTensor function for performing the computation
f = _backend.function(inputs, outputs)

# compute the model outcome
actual = f(0.85, 2.0)

assert numpy.allclose(actual[0], [2.0, 1.4819299, 1.28322046])
assert numpy.allclose(actual[1], [0.71677954, 0.83004323])
assert numpy.allclose(actual[2], [0.5180701, 0.71677954, 0.83004323])
return

@pytest.mark.skipif(not HAS_PYMC, reason="requires PyMC")
Expand Down
Loading