vivarium-public-health 3.0.4__tar.gz → 3.0.7__tar.gz

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/CHANGELOG.rst

@@ -1,3 +1,19 @@
+**3.0.7 - 09/10/24**
+
+ - Properly create the population_view in RateTransition
+
+**3.0.6 - 09/04/24**
+
+ - Fix bug that was occurring when RiskEffect's rr_source was a float or DataFrame
+ - Better handle exposure column creation in Risk component
+ - Rename LBWSRisk 'exposure_column_name()' staticmethod to not collide with Risk attr
+
+**3.0.5 - 08/29/24**
+
+ - Add missing results module-level docstrings
+ - Strengthen the SimpleCause docstring
+ - Rename the SimpleCause class method
+
 **3.0.4 - 08/28/24**
 
  - Strengthen results docstrings and clean up others throughout

{vivarium_public_health-3.0.4/src/vivarium_public_health.egg-info → vivarium_public_health-3.0.7}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vivarium_public_health
-Version: 3.0.4
+Version: 3.0.7
 Summary: Components for modelling diseases, risks, and interventions with ``vivarium``
 Home-page: https://github.com/ihmeuw/vivarium_public_health
 Author: The vivarium developers

vivarium_public_health-3.0.7/src/vivarium_public_health/_version.py

@@ -0,0 +1 @@
+__version__ = "3.0.7"

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/src/vivarium_public_health/disease/transition.py

@@ -7,7 +7,7 @@ This module contains tools to model transitions between disease states.
 
 """
 
-from typing import TYPE_CHECKING, Any, Callable, Dict, Union
+from typing import TYPE_CHECKING, Any, Callable, Dict, List, Optional, Union
 
 import pandas as pd
 from vivarium.framework.engine import Builder
@@ -48,6 +48,10 @@ class RateTransition(Transition):
             },
         }
 
+    @property
+    def columns_required(self) -> Optional[List[str]]:
+        return ["alive"]
+
     @property
     def transition_rate_pipeline_name(self) -> str:
         if "incidence_rate" in self._get_data_functions:
@@ -100,8 +104,6 @@ class RateTransition(Transition):
             preferred_post_processor=union_post_processor,
         )
 
-        self.population_view = builder.population.get_view(["alive"])
-
     #################
     # Setup methods #
     #################

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/src/vivarium_public_health/results/disability.py

@@ -1,7 +1,7 @@
 """
-===================
-Disability Observer
-===================
+====================
+Disability Observers
+====================
 
 This module contains tools for observing years lived with disability (YLDs)
 in the simulation.
@@ -102,7 +102,7 @@ class DisabilityObserver(PublicHealthObserver):
         )
         # Convert to SimpleCause instances and add on all_causes
         causes_of_disability = [
-            SimpleCause.create_from_disease_state(cause) for cause in causes_of_disability
+            SimpleCause.create_from_specific_cause(cause) for cause in causes_of_disability
         ] + [SimpleCause("all_causes", "all_causes", "cause")]
 
         excluded_causes = (

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/src/vivarium_public_health/results/disease.py

@@ -1,7 +1,7 @@
 """
-================
-Disease Observer
-================
+=================
+Disease Observers
+=================
 
 This module contains tools for observing disease incidence and prevalence
 in the simulation.

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/src/vivarium_public_health/results/mortality.py

@@ -1,7 +1,7 @@
 """
-==================
-Mortality Observer
-==================
+===================
+Mortality Observers
+===================
 
 This module contains tools for observing cause-specific and
 excess mortality in the simulation, including "other causes".
@@ -130,7 +130,7 @@ class MortalityObserver(PublicHealthObserver):
 
         # Convert to SimpleCauses and add on other_causes and not_dead
         self.causes_of_death = [
-            SimpleCause.create_from_disease_state(cause) for cause in causes_of_death
+            SimpleCause.create_from_specific_cause(cause) for cause in causes_of_death
         ] + [
             SimpleCause("not_dead", "not_dead", "cause"),
             SimpleCause("other_causes", "other_causes", "cause"),

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/src/vivarium_public_health/results/observer.py

@@ -1,3 +1,13 @@
+"""
+===============
+Basic Observers
+===============
+
+This module contains convenience classes for building concrete observers in
+public health models.
+
+"""
+
 from typing import Callable, List, Optional, Union
 
 import pandas as pd

vivarium_public_health-3.0.7/src/vivarium_public_health/results/simple_cause.py

@@ -0,0 +1,43 @@
+"""
+============
+Simple Cause
+============
+
+This module contains tools for creating a minimal representation of a cause
+as required by observers.
+
+"""
+
+from dataclasses import dataclass
+
+
+@dataclass
+class SimpleCause:
+    """A simple dataclass to represent the bare minimum information needed by observers.
+
+    It also includes a class method to convert a provided cause into a
+    ``SimpleCause`` instance.
+
+    """
+
+    state_id: str
+    """The state_id of the cause."""
+    model: str
+    """The model of the cause."""
+    cause_type: str
+    """The cause type of the cause."""
+
+    @classmethod
+    def create_from_specific_cause(cls, cause: type) -> "SimpleCause":
+        """Create a SimpleCause instance from a more specific cause.
+
+        Parameters
+        ----------
+        cause
+            The cause to be converted into a SimpleCause instance.
+
+        Returns
+        -------
+            A SimpleCause instance.
+        """
+        return cls(cause.state_id, cause.model, cause.cause_type)

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/src/vivarium_public_health/risks/base_risk.py

@@ -123,7 +123,10 @@ class Risk(Component):
 
     @property
     def columns_created(self) -> List[str]:
-        return [self.propensity_column_name, self.exposure_column_name]
+        columns_to_create = [self.propensity_column_name]
+        if self.create_exposure_column:
+            columns_to_create.append(self.exposure_column_name)
+        return columns_to_create
 
     @property
     def initialization_requirements(self) -> Dict[str, List[str]]:
@@ -172,6 +175,16 @@ class Risk(Component):
         self.propensity = self.get_propensity_pipeline(builder)
         self.exposure = self.get_exposure_pipeline(builder)
 
+        # We want to set this to True iff there is a non-loglinear risk effect
+        # on this risk instance
+        self.create_exposure_column = bool(
+            [
+                component
+                for component in builder.components.list_components()
+                if component.startswith(f"non_log_linear_risk_effect.{self.risk.name}_on_")
+            ]
+        )
+
     def get_distribution_type(self, builder: Builder) -> str:
         """Get the distribution type for the risk from the configuration.
 
@@ -270,19 +283,19 @@ class Risk(Component):
     ########################
 
     def on_initialize_simulants(self, pop_data: SimulantData) -> None:
-        propensity_values = self.randomness.get_draw(pop_data.index)
-        df = pd.DataFrame(
-            {
-                self.propensity_column_name: self.randomness.get_draw(pop_data.index),
-                self.exposure_column_name: self.exposure_distribution.ppf(propensity_values),
-            }
+        propensity = pd.Series(
+            self.randomness.get_draw(pop_data.index), name=self.propensity_column_name
         )
-        self.population_view.update(df)
+        self.population_view.update(propensity)
+        self.update_exposure_column(pop_data.index)
 
     def on_time_step_prepare(self, event: Event) -> None:
-        exposure_values = self.exposure(event.index)
-        exposure_col = pd.Series(exposure_values, name=self.exposure_column_name)
-        self.population_view.update(exposure_col)
+        self.update_exposure_column(event.index)
+
+    def update_exposure_column(self, index: pd.Index) -> None:
+        if self.create_exposure_column:
+            exposure = pd.Series(self.exposure(index), name=self.exposure_column_name)
+            self.population_view.update(exposure)
 
     ##################################
     # Pipeline sources and modifiers #

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/src/vivarium_public_health/risks/effect.py

@@ -157,16 +157,19 @@ class RiskEffect(Component):
         rr_source = configuration.data_sources.relative_risk
         rr_dist_parameters = configuration.data_source_parameters.relative_risk.to_dict()
 
-        try:
-            distribution = getattr(import_module("scipy.stats"), rr_source)
-            rng = np.random.default_rng(builder.randomness.get_seed(self.name))
-            rr_data = distribution(**rr_dist_parameters).ppf(rng.random())
-        except AttributeError:
+        if isinstance(rr_source, str):
+            try:
+                distribution = getattr(import_module("scipy.stats"), rr_source)
+                rng = np.random.default_rng(builder.randomness.get_seed(self.name))
+                rr_data = distribution(**rr_dist_parameters).ppf(rng.random())
+            except AttributeError:
+                rr_data = self.get_filtered_data(builder, rr_source)
+            except TypeError:
+                raise ConfigurationError(
+                    f"Parameters {rr_dist_parameters} are not valid for distribution {rr_source}."
+                )
+        else:
             rr_data = self.get_filtered_data(builder, rr_source)
-        except TypeError:
-            raise ConfigurationError(
-                f"Parameters {rr_dist_parameters} are not valid for distribution {rr_source}."
-            )
         return rr_data
 
     def get_filtered_data(
@@ -198,6 +201,8 @@ class RiskEffect(Component):
             cat2["parameter"] = "cat2"
             cat2["value"] = 1
             rr_data = pd.concat([cat1, cat2], ignore_index=True)
+        if "parameter" in rr_data.index.names:
+            rr_data = rr_data.reset_index("parameter")
 
         rr_value_cols = list(rr_data["parameter"].unique())
         rr_data = pivot_categorical(builder, self.risk, rr_data, "parameter")
@@ -361,6 +366,10 @@ class NonLogLinearRiskEffect(RiskEffect):
     # Setup methods #
     #################
 
+    @staticmethod
+    def get_name(risk: EntityString, target: TargetString) -> str:
+        return f"non_log_linear_risk_effect.{risk.name}_on_{target}"
+
     def build_all_lookup_tables(self, builder: Builder) -> None:
         rr_data = self.get_relative_risk_data(builder)
         self.validate_rr_data(rr_data)

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/src/vivarium_public_health/risks/implementations/low_birth_weight_and_short_gestation.py

@@ -191,7 +191,7 @@ class LBWSGRisk(Risk):
         return f"{axis}.birth_exposure"
 
     @staticmethod
-    def exposure_column_name(axis: str) -> str:
+    def get_exposure_column_name(axis: str) -> str:
         return f"{axis}_exposure"
 
     ##############
@@ -206,7 +206,7 @@ class LBWSGRisk(Risk):
 
     @property
     def columns_created(self) -> List[str]:
-        return [self.exposure_column_name(axis) for axis in self.AXES]
+        return [self.get_exposure_column_name(axis) for axis in self.AXES]
 
     #####################
     # Lifecycle methods #
@@ -256,7 +256,7 @@ class LBWSGRisk(Risk):
 
     def on_initialize_simulants(self, pop_data: SimulantData) -> None:
         birth_exposures = {
-            self.exposure_column_name(axis): self.birth_exposures[
+            self.get_exposure_column_name(axis): self.birth_exposures[
                 self.birth_exposure_pipeline_name(axis)
             ](pop_data.index)
             for axis in self.AXES
@@ -318,7 +318,7 @@ class LBWSGRiskEffect(RiskEffect):
         super().__init__("risk_factor.low_birth_weight_and_short_gestation", target)
 
         self.lbwsg_exposure_column_names = [
-            LBWSGRisk.exposure_column_name(axis) for axis in LBWSGRisk.AXES
+            LBWSGRisk.get_exposure_column_name(axis) for axis in LBWSGRisk.AXES
         ]
         self.relative_risk_pipeline_name = (
             f"effect_of_{self.risk.name}_on_{self.target.name}.relative_risk"
@@ -433,8 +433,8 @@ class LBWSGRiskEffect(RiskEffect):
         pop = self.population_view.subview(["sex"] + self.lbwsg_exposure_column_names).get(
             pop_data.index
         )
-        birth_weight = pop[LBWSGRisk.exposure_column_name(BIRTH_WEIGHT)]
-        gestational_age = pop[LBWSGRisk.exposure_column_name(GESTATIONAL_AGE)]
+        birth_weight = pop[LBWSGRisk.get_exposure_column_name(BIRTH_WEIGHT)]
+        gestational_age = pop[LBWSGRisk.get_exposure_column_name(GESTATIONAL_AGE)]
 
         is_male = pop["sex"] == "Male"
         is_tmrel = (self.TMREL_GESTATIONAL_AGE_INTERVAL.left <= gestational_age) & (

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7/src/vivarium_public_health.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: vivarium_public_health
-Version: 3.0.4
+Version: 3.0.7
 Summary: Components for modelling diseases, risks, and interventions with ``vivarium``
 Home-page: https://github.com/ihmeuw/vivarium_public_health
 Author: The vivarium developers

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/tests/mock_artifact.py

@@ -77,6 +77,7 @@ MOCKERS = {
     "population": {
         "age_bins": make_age_bins(),
         "structure": make_uniform_pop_data(),
+        "demographic_dimensions": make_uniform_pop_data().drop(["location", "value"], axis=1),
         "theoretical_minimum_risk_life_expectancy": (
             build_table_with_age(
                 98.0,

{vivarium_public_health-3.0.4 → vivarium_public_health-3.0.7}/tests/risks/test_effect.py

@@ -23,6 +23,7 @@ from vivarium_public_health.risks.base_risk import Risk
 from vivarium_public_health.risks.effect import NonLogLinearRiskEffect, RiskEffect
 from vivarium_public_health.utilities import EntityString
 
+
 #
 #
 # def test_incidence_rate_risk_effect(base_config, base_plugins, mocker):
@@ -413,6 +414,95 @@ from vivarium_public_health.utilities import EntityString
 #                                                   source=lambda index: pd.Series(0.1, index=index))
 #
 #     assert np.allclose(from_yearly(0.1, time_step)*50, em(simulation.get_population().index))
+def _setup_risk_effect_simulation(
+    config: LayeredConfigTree,
+    plugins: LayeredConfigTree,
+    risk: Union[str, Risk],
+    risk_effect: RiskEffect,
+    data: Dict[str, Any],
+) -> InteractiveContext:
+    components = [
+        TestPopulation(),
+        risk,
+        SI("test_cause"),
+        risk_effect,
+    ]
+
+    simulation = InteractiveContext(
+        components=components,
+        configuration=config,
+        plugin_configuration=plugins,
+        setup=False,
+    )
+
+    for key, value in data.items():
+        simulation._data.write(key, value)
+
+    simulation.setup()
+    return simulation
+
+
+def build_dichotomous_risk_effect_data(rr_value: float) -> pd.DataFrame:
+    df = pd.DataFrame(
+        {
+            "affected_entity": "test_cause",
+            "affected_measure": "incidence_rate",
+            "year_start": 1990,
+            "year_end": 1991,
+            "value": [rr_value, 1.0],
+        },
+        index=pd.Index(["cat1", "cat2"], name="parameter"),
+    )
+    return df
+
+
+@pytest.mark.parametrize(
+    "rr_source, rr_value",
+    [("str", 2.0), ("float", 0.9), ("DataFrame", 0.5)],
+)
+def test_rr_sources(rr_source, rr_value, dichotomous_risk, base_config, base_plugins):
+    risk = dichotomous_risk[0]
+    effect = RiskEffect(risk.name, "cause.test_cause.incidence_rate")
+    base_config.update({"risk_factor.test_risk": {"data_sources": {"exposure": 1.0}}})
+
+    # TMREL of 1
+    tmred = {"distribution": "uniform", "min": 1, "max": 1, "inverted": False}
+
+    data = {
+        f"{risk.name}.tmred": tmred,
+        f"{risk.name}.population_attributable_fraction": 0,
+        "cause.test_cause.incidence_rate": 1,
+    }
+
+    if rr_source == "DataFrame":
+        rr_data = build_dichotomous_risk_effect_data(rr_value)
+        base_config.update(
+            {
+                "risk_effect.test_risk_on_cause.test_cause.incidence_rate": {
+                    "data_sources": {"relative_risk": rr_data}
+                }
+            }
+        )
+    elif rr_source == "float":
+        base_config.update(
+            {
+                "risk_effect.test_risk_on_cause.test_cause.incidence_rate": {
+                    "data_sources": {"relative_risk": rr_value}
+                }
+            }
+        )
+    else:  # rr_source is a string because it gets read from RiskEffect's configuration defaults
+        rr_data = build_dichotomous_risk_effect_data(rr_value)
+        data[f"{risk.name}.relative_risk"] = rr_data
+
+    base_config.update({"risk_factor.test_risk": {"distribution_type": "dichotomous"}})
+    simulation = _setup_risk_effect_simulation(base_config, base_plugins, risk, effect, data)
+
+    pop = simulation.get_population()
+    rate = simulation.get_value("test_cause.incidence_rate")(
+        pop.index, skip_post_processor=True
+    )
+    assert set(rate.unique()) == {rr_value}
 
 
 ##############################
@@ -458,33 +548,6 @@ class CustomExposureRisk(Component):
         return data
 
 
-def _setup_risk_simulation(
-    config: LayeredConfigTree,
-    plugins: LayeredConfigTree,
-    risk: Union[str, Risk],
-    data: Dict[str, Any],
-) -> InteractiveContext:
-    components = [
-        TestPopulation(),
-        risk,
-        SI("some_disease"),
-        NonLogLinearRiskEffect(risk.name, "cause.some_disease.incidence_rate"),
-    ]
-
-    simulation = InteractiveContext(
-        components=components,
-        configuration=config,
-        plugin_configuration=plugins,
-        setup=False,
-    )
-
-    for key, value in data.items():
-        simulation._data.write(key, value)
-
-    simulation.setup()
-    return simulation
-
-
 @pytest.mark.parametrize(
     "rr_parameter_data, error_message",
     [
@@ -496,14 +559,12 @@ def _setup_risk_simulation(
 )
 def test_non_loglinear_effect(rr_parameter_data, error_message, base_config, base_plugins):
     risk = CustomExposureRisk("risk_factor.test_risk")
-    effect = NonLogLinearRiskEffect(
-        "risk_factor.test_risk", "cause.some_disease.incidence_rate"
-    )
+    effect = NonLogLinearRiskEffect(risk.name, "cause.test_cause.incidence_rate")
 
     risk_effect_rrs = [2.0, 2.4, 4.0]
     rr_data = pd.DataFrame(
         {
-            "affected_entity": "some_disease",
+            "affected_entity": "test_cause",
             "affected_measure": "incidence_rate",
             "year_start": 1990,
             "year_end": 1991,
@@ -518,20 +579,24 @@ def test_non_loglinear_effect(rr_parameter_data, error_message, base_config, bas
         f"{risk.name}.relative_risk": rr_data,
         f"{risk.name}.tmred": tmred,
         f"{risk.name}.population_attributable_fraction": 0,
-        "cause.some_disease.incidence_rate": 1,
+        "cause.test_cause.incidence_rate": 1,
     }
 
     base_config.update({"population": {"population_size": 10}})
 
     if error_message:
         with pytest.raises(ValueError, match=error_message):
-            simulation = _setup_risk_simulation(base_config, base_plugins, risk, data)
+            simulation = _setup_risk_effect_simulation(
+                base_config, base_plugins, risk, effect, data
+            )
         return
     else:
-        simulation = _setup_risk_simulation(base_config, base_plugins, risk, data)
+        simulation = _setup_risk_effect_simulation(
+            base_config, base_plugins, risk, effect, data
+        )
 
     pop = simulation.get_population()
-    rate = simulation.get_value("some_disease.incidence_rate")(
+    rate = simulation.get_value("test_cause.incidence_rate")(
         pop.index, skip_post_processor=True
     )
     expected_values = np.interp(

vivarium_public_health-3.0.4/src/vivarium_public_health/_version.py

@@ -1 +0,0 @@
-__version__ = "3.0.4"

vivarium_public_health-3.0.4/src/vivarium_public_health/results/simple_cause.py

@@ -1,24 +0,0 @@
-from dataclasses import dataclass
-
-
-@dataclass
-class SimpleCause:
-    """A simple dataclass to represent the bare minimum information needed
-    for observers, e.g. 'all_causes' as a cause of disability.
-
-    It also includes a class method to convert a provided disease state into a
-    ``SimpleCause`` instance.
-
-    """
-
-    state_id: str
-    """The state_id of the cause."""
-    model: str
-    """The model of the cause."""
-    cause_type: str
-    """The cause type of the cause."""
-
-    @classmethod
-    def create_from_disease_state(cls, disease_state: type) -> "SimpleCause":
-        """Create a SimpleCause instance from a"""
-        return cls(disease_state.state_id, disease_state.model, disease_state.cause_type)