Merge pull request #320 from ihmeuw/develop

releave v0.11.0

.github/workflows/build.yml

@@ -10,7 +10,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: ["3.7", "3.8", "3.9", "3.10"]
+        python-version: ["3.8", "3.9", "3.10", "3.11"]
     defaults:
       run:
         shell: bash -le {0}

.github/workflows/deploy.yml

@@ -12,7 +12,7 @@ jobs:
       - name: Set up Python
         uses: actions/setup-python@v4
         with:
-          python-version: '3.8'
+          python-version: '3.11'
       - name: Install dependencies
         run: |
           python --version

CHANGELOG.rst

@@ -1,3 +1,10 @@
+**0.11.0 - 06/01/23**
+
+ - Support Python 3.8-3.11
+ - Update vivarium pin
+ - Handle FutureWarning
+ - Refactor to create a 'get_transition_names' function
+
 **0.10.24 - 05/11/23**
 
  - Standardize builder, cause argument order in state get data functions

setup.py

@@ -4,7 +4,6 @@
 from setuptools import find_packages, setup
 
 if __name__ == "__main__":
-
     base_dir = os.path.dirname(__file__)
     src_dir = os.path.join(base_dir, "src")
 
@@ -16,7 +15,7 @@
         long_description = f.read()
 
     install_requirements = [
-        "vivarium>=1.1.0",
+        "vivarium>=1.2.0",
         "numpy",
         "pandas",
         "scipy",

src/vivarium_public_health/__about__.py

@@ -13,7 +13,7 @@
 __summary__ = "Components for modelling diseases, risks, and interventions with ``vivarium``"
 __uri__ = "https://github.com/ihmeuw/vivarium_public_health"
 
-__version__ = "0.10.24"
+__version__ = "0.11.0"
 
 __author__ = "The vivarium_public_health developers"
 __email__ = "[email protected]"

src/vivarium_public_health/disease/model.py

@@ -53,9 +53,7 @@ def transition_names(self) -> List[TransitionString]:
         states = {s.name.split(".")[1]: s for s in self.states}
         transitions = []
         for state in states.values():
-            for trans in state.transition_set.transitions:
-                _, _, init_state, _, end_state = trans.name.split(".")
-                transitions.append(TransitionString(f"{init_state}_TO_{end_state}"))
+            transitions += state.get_transition_names()
         return transitions
 
     def setup(self, builder):

src/vivarium_public_health/disease/special_disease.py

@@ -283,7 +283,6 @@ def with_condition(self, index):
         return with_condition
 
     def get_exposure_filter(self, distribution, exposure_pipeline, threshold):
-
         if distribution in ["dichotomous", "ordered_polytomous", "unordered_polytomous"]:
 
             def categorical_filter(index):

src/vivarium_public_health/disease/state.py

@@ -6,7 +6,7 @@
 This module contains tools to manage standard disease states.
 
 """
-from typing import Callable, Dict
+from typing import Callable, Dict, List
 
 import numpy as np
 import pandas as pd
@@ -17,6 +17,7 @@
 from vivarium_public_health.disease.transition import (
     ProportionTransition,
     RateTransition,
+    TransitionString,
 )
 from vivarium_public_health.utilities import is_non_zero
 
@@ -94,6 +95,17 @@ def _transition_side_effect(self, index, event_time):
         if self.side_effect_function is not None:
             self.side_effect_function(index, event_time)
 
+    ##################
+    # Public methods #
+    ##################
+
+    def get_transition_names(self) -> List[str]:
+        transitions = []
+        for trans in self.transition_set.transitions:
+            _, _, init_state, _, end_state = trans.name.split(".")
+            transitions.append(TransitionString(f"{init_state}_TO_{end_state}"))
+        return transitions
+
     def add_transition(
         self,
         output: State,

src/vivarium_public_health/population/data_transformations.py

@@ -108,13 +108,13 @@ def rescale_binned_proportions(
         values are rescaled to reflect their smaller representation.
     """
     col_order = pop_data.columns.copy()
-    if age_start > pop_data.age_end.max():
+    if age_start > pop_data["age_end"].max():
         raise ValueError(
             "Provided population data is insufficient to model the requested age range."
         )
 
-    age_start = max(pop_data.age_start.min(), age_start)
-    age_end = min(pop_data.age_end.max(), age_end) - 1e-8
+    age_start = max(pop_data["age_start"].min(), age_start)
+    age_end = min(pop_data["age_end"].max(), age_end) - 1e-8
     pop_data = _add_edge_age_groups(pop_data.copy())
 
     columns_to_scale = [
@@ -123,12 +123,13 @@ def rescale_binned_proportions(
         "value",
     ]
     for _, sub_pop in pop_data.groupby(["sex", "location"]):
+        min_bin = sub_pop[
+            (sub_pop["age_start"] <= age_start) & (age_start < sub_pop["age_end"])
+        ]
+        padding_bin = sub_pop[sub_pop["age_end"] == float(min_bin["age_start"].iloc[0])]
 
-        min_bin = sub_pop[(sub_pop.age_start <= age_start) & (age_start < sub_pop.age_end)]
-        padding_bin = sub_pop[sub_pop.age_end == float(min_bin.age_start)]
-
-        min_scale = (float(min_bin.age_end) - age_start) / float(
-            min_bin.age_end - min_bin.age_start
+        min_scale = (float(min_bin["age_end"].iloc[0]) - age_start) / float(
+            min_bin["age_end"].iloc[0] - min_bin["age_start"].iloc[0]
         )
 
         remainder = pop_data.loc[min_bin.index, columns_to_scale].values * (1 - min_scale)
@@ -138,11 +139,11 @@ def rescale_binned_proportions(
         pop_data.loc[min_bin.index, "age_start"] = age_start
         pop_data.loc[padding_bin.index, "age_end"] = age_start
 
-        max_bin = sub_pop[(sub_pop.age_end > age_end) & (age_end >= sub_pop.age_start)]
-        padding_bin = sub_pop[sub_pop.age_start == float(max_bin.age_end)]
+        max_bin = sub_pop[(sub_pop["age_end"] > age_end) & (age_end >= sub_pop["age_start"])]
+        padding_bin = sub_pop[sub_pop["age_start"] == float(max_bin["age_end"].iloc[0])]
 
-        max_scale = (age_end - float(max_bin.age_start)) / float(
-            max_bin.age_end - max_bin.age_start
+        max_scale = (age_end - float(max_bin["age_start"].iloc[0])) / float(
+            max_bin["age_end"].iloc[0] - max_bin["age_start"].iloc[0]
         )
 
         remainder = pop_data.loc[max_bin.index, columns_to_scale] * (1 - max_scale)
@@ -239,20 +240,21 @@ def smooth_ages(
     """
     simulants = simulants.copy()
     for (sex, location), sub_pop in population_data.groupby(["sex", "location"]):
-
-        ages = sorted(sub_pop.age.unique())
-        younger = [float(sub_pop.loc[sub_pop.age == ages[0], "age_start"])] + ages[:-1]
-        older = ages[1:] + [float(sub_pop.loc[sub_pop.age == ages[-1], "age_end"])]
+        ages = sorted(sub_pop["age"].unique())
+        younger = [float(sub_pop.loc[sub_pop["age"] == ages[0], "age_start"].iloc[0])] + ages[
+            :-1
+        ]
+        older = ages[1:] + [float(sub_pop.loc[sub_pop["age"] == ages[-1], "age_end"].iloc[0])]
 
         uniform_all = randomness.get_draw(simulants.index)
 
         for age_set in zip(ages, younger, older):
             age = AgeValues(*age_set)
 
             has_correct_demography = (
-                (simulants.age == age.current)
-                & (simulants.sex == sex)
-                & (simulants.location == location)
+                (simulants["age"] == age.current)
+                & (simulants["sex"] == sex)
+                & (simulants["location"] == location)
             )
             affected = simulants[has_correct_demography]
 
@@ -319,15 +321,15 @@ def _get_bins_and_proportions(
         )
 
     """
-    left = float(pop_data.loc[pop_data.age == age.current, "age_start"])
-    right = float(pop_data.loc[pop_data.age == age.current, "age_end"])
+    left = float(pop_data.loc[pop_data["age"] == age.current, "age_start"].iloc[0])
+    right = float(pop_data.loc[pop_data["age"] == age.current, "age_end"].iloc[0])
 
-    if not pop_data.loc[pop_data.age == age.young, "age_start"].empty:
-        lower_left = float(pop_data.loc[pop_data.age == age.young, "age_start"])
+    if not pop_data.loc[pop_data["age"] == age.young, "age_start"].empty:
+        lower_left = float(pop_data.loc[pop_data["age"] == age.young, "age_start"].iloc[0])
     else:
         lower_left = left
-    if not pop_data.loc[pop_data.age == age.old, "age_end"].empty:
-        upper_right = float(pop_data.loc[pop_data.age == age.old, "age_end"])
+    if not pop_data.loc[pop_data["age"] == age.old, "age_end"].empty:
+        upper_right = float(pop_data.loc[pop_data["age"] == age.old, "age_end"].iloc[0])
     else:
         upper_right = right
 
@@ -338,18 +340,21 @@ def _get_bins_and_proportions(
     # in order to back out a point estimate for the probability density at the center
     # of the interval. This not the best assumption, but it'll do.
     p_age = float(
-        pop_data.loc[pop_data.age == age.current, proportion_column] / (right - left)
+        pop_data.loc[pop_data["age"] == age.current, proportion_column].iloc[0]
+        / (right - left)
     )
     p_young = (
         float(
-            pop_data.loc[pop_data.age == age.young, proportion_column] / (left - lower_left)
+            pop_data.loc[pop_data["age"] == age.young, proportion_column].iloc[0]
+            / (left - lower_left)
         )
         if age.young != left
         else p_age
     )
     p_old = (
         float(
-            pop_data.loc[pop_data.age == age.old, proportion_column] / (upper_right - right)
+            pop_data.loc[pop_data["age"] == age.old, proportion_column].iloc[0]
+            / (upper_right - right)
         )
         if age.old != right
         else 0

src/vivarium_public_health/population/mortality.py

@@ -56,7 +56,6 @@
 
 
 class Mortality:
-
     configuration_defaults = {"unmodeled_causes": []}
 
     def __init__(self):

src/vivarium_public_health/risks/implementations/low_birth_weight_and_short_gestation.py

@@ -31,7 +31,6 @@
 
 
 class LBWSGDistribution(PolytomousDistribution):
-
     configuration_defaults = {
         "lbwsg_distribution": {
             "age_column": "age",
@@ -211,7 +210,6 @@ def _parse_description(axis: str, description: str) -> pd.Interval:
 
 
 class LBWSGRisk(Risk):
-
     AXES = [BIRTH_WEIGHT, GESTATIONAL_AGE]
 
     def __init__(self):
@@ -316,7 +314,6 @@ def _get_current_exposure(self, index: pd.Index) -> pd.DataFrame:
 
 
 class LBWSGRiskEffect(RiskEffect):
-
     TMREL_BIRTH_WEIGHT_INTERVAL: pd.Interval = pd.Interval(3500.0, 4500.0)
     TMREL_GESTATIONAL_AGE_INTERVAL: pd.Interval = pd.Interval(38.0, 42.0)
 

tests/disease/test_disease.py

@@ -52,7 +52,6 @@ def get_test_prevalence(simulation, key):
 
 
 def test_dwell_time(assign_cause_mock, base_config, base_plugins, disease, base_data):
-
     time_step = 10
     assign_cause_mock.side_effect = lambda population, *args: pd.DataFrame(
         {"condition_state": "healthy"}, index=population.index

tests/metrics/test_stratification.py

@@ -15,7 +15,6 @@
 
 
 class FavoriteColor:
-
     OPTIONS = ["red", "green", "orange"]
 
     @property
@@ -34,7 +33,6 @@ def setup(self, builder: Builder):
 
 
 class FavoriteNumber:
-
     OPTIONS = [7, 42, 14312]
 
     @property

tests/population/test_add_new_birth_cohort.py

@@ -16,7 +16,6 @@
 
 @pytest.fixture()
 def config(base_config):
-
     base_config.update(
         {
             "population": {