Merge pull request #176 from CodeForPhilly/move_parameters

Move parameters

Author: quinn-dougherty

src/app.py

@@ -1,3 +1,5 @@
+"""App."""
+
 import altair as alt
 import streamlit as st
 

src/cli.py

@@ -1,4 +1,4 @@
-"""CLI for CHIME."""
+"""Command line interface."""
 
 from argparse import (
     Action,
@@ -8,9 +8,8 @@
 
 from pandas import DataFrame
 
-from penn_chime.defaults import RateLos
-from penn_chime.models import Parameters
-from penn_chime.utils import build_admissions_df, build_census_df
+from penn_chime.parameters import Parameters
+from penn_chime.utils import build_admissions_df, build_census_df, RateLos
 
 
 class FromFile(Action):

src/penn_chime/defaults.py

@@ -1,14 +1,6 @@
-#!/usr/bin/env python
-"""Set defaults for your fork/locality here
+"""Defaults."""
 
-   after we merge this in we can set defaults in a `config/env` type of file
-"""
-
-from collections import namedtuple
-
-# (0.02, 7) is 2%, 7 days
-# be sure to multiply by 100 when using as a default to the pct widgets!
-RateLos = namedtuple('RateLos', ('rate', 'length_of_stay'))
+from .utils import RateLos
 
 
 class Regions:
@@ -54,4 +46,3 @@ def __init__(
 
     def __repr__(self) -> str:
         return f"Constants(susceptible_default: {self.region.susceptible}, known_infected: {self.known_infected})"
-

src/penn_chime/models.py

@@ -1,122 +1,11 @@
+"""Models."""
+
 from typing import Generator, Tuple
 
 import numpy as np
 import pandas as pd
-import streamlit as st
-
-from .defaults import RateLos
-
-
-class Parameters:
-
-    def __init__(
-        self, *,
-        current_hospitalized: int,
-        doubling_time: float,
-        known_infected: int,
-        market_share: float,
-        relative_contact_rate: float,
-        susceptible: int,
-
-        hospitalized: RateLos,
-        icu: RateLos,
-        ventilated: RateLos,
-
-        max_y_axis: int = None,
-        n_days: int = None
-    ):
-        self.current_hospitalized = current_hospitalized
-        self.doubling_time = doubling_time
-        self.known_infected = known_infected
-        self.market_share = market_share
-        self.relative_contact_rate = relative_contact_rate
-        self.susceptible = susceptible
-        self._n_days = 0
-
-        self.hospitalized = hospitalized
-        self.icu = icu
-        self.ventilated = ventilated
-
-        self.max_y_axis = max_y_axis
-
-        self.rates = tuple(
-            each.rate
-            for each in (hospitalized, icu, ventilated)
-        )
-        self.lengths_of_stay = tuple(
-            each.length_of_stay
-            for each in (hospitalized, icu, ventilated)
-        )
-
-        # Note: this should not be an integer.
-        # We're appoximating infected from what we do know.
-        # TODO market_share > 0, hosp_rate > 0
-        self.infected = infected = \
-            current_hospitalized / market_share / hospitalized.rate
-
-        self.detection_probability = \
-            known_infected / infected if infected > 1.e-7 else None
-
-        # TODO missing initial recovered value
-        self.recovered = 0.0
-
-        self.intrinsic_growth_rate = intrinsic_growth_rate = \
-            2.0 ** (1.0 / doubling_time) - 1.0 if doubling_time > 0.0 else 0.0
-
-        # TODO make this configurable, or more nuanced
-        self.recovery_days = recovery_days = 14.0
-
-        self.gamma = gamma = 1.0 / recovery_days
-
-        # Contact rate, beta
-        self.beta = beta = (
-            (intrinsic_growth_rate + gamma) /
-            susceptible * (1.0 - relative_contact_rate)
-        )  # {rate based on doubling time} / {initial susceptible}
-
-        # r_t is r_0 after distancing
-        self.r_t = beta / gamma * susceptible
-
-        # Simplify equation to avoid division by zero:
-        # self.r_naught = r_t / (1.0 - relative_contact_rate)
-        self.r_naught = (intrinsic_growth_rate + gamma) / gamma
-
-        # doubling time after distancing
-        # TODO constrain values np.log2(...) > 0.0
-        self.doubling_time_t = 1.0 / np.log2(beta * susceptible - gamma + 1)
-
-        self.dispositions = None
-        self.susceptible_v = self.infected_v = self.recovered_v = None
-        self.hospitalized_v = self.icu_v = self.ventilated_v = None
-
-        if n_days is not None:
-            self.n_days = n_days
-
-    @property
-    def n_days(self):
-        return self._n_days
-
-    @n_days.setter
-    def n_days(self, n_days: int):
-        self._n_days = n_days
-
-        s_v, i_v, r_v = sim_sir(
-            self.susceptible,
-            self.infected,
-            self.recovered,
-            self.beta,
-            self.gamma,
-            n_days,
-        )
-        self.susceptible_v, self.infected_v, self.recovered_v = s_v, i_v, r_v
-
-        self.dispositions = hospitalized_v, icu_v, ventilated_v = \
-            get_dispositions(i_v, self.rates, self.market_share)
-        self.hospitalized_v, self.icu_v, self.ventilated_v = \
-            hospitalized_v, icu_v, ventilated_v
 
 
-@st.cache
 def sir(
     s: float, i: float, r: float,
     beta: float, gamma: float, n: float
@@ -148,7 +37,6 @@ def gen_sir(
         s, i, r = sir(s, i, r, beta, gamma, n)
 
 
-@st.cache
 def sim_sir(
     s: float, i: float, r: float,
     beta: float, gamma: float, n_days: int
@@ -170,7 +58,6 @@ def sim_sir(
     )
 
 
-@st.cache
 def sim_sir_df(
     s: float, i: float, r: float,
     beta: float, gamma: float, n_days: int
@@ -182,7 +69,6 @@ def sim_sir_df(
     )
 
 
-@st.cache
 def get_dispositions(
     infected: np.ndarray, rates: Tuple[float, ...], market_share: float = 1.0
 ) -> Tuple[np.ndarray, ...]:

src/penn_chime/parameters.py

@@ -0,0 +1,118 @@
+"""Parameters."""
+
+from numpy import log2
+
+from .utils import RateLos
+from .models import (
+    get_dispositions,
+    sim_sir,
+)
+
+
+class Parameters:
+
+    def __init__(
+        self, *,
+        current_hospitalized: int,
+        doubling_time: float,
+        known_infected: int,
+        market_share: float,
+        relative_contact_rate: float,
+        susceptible: int,
+
+        hospitalized: RateLos,
+        icu: RateLos,
+        ventilated: RateLos,
+
+        max_y_axis: int = None,
+        n_days: int = None
+    ):
+        self.current_hospitalized = current_hospitalized
+        self.doubling_time = doubling_time
+        self.known_infected = known_infected
+        self.market_share = market_share
+        self.relative_contact_rate = relative_contact_rate
+        self.susceptible = susceptible
+        self._n_days = 0
+
+        self.hospitalized = hospitalized
+        self.icu = icu
+        self.ventilated = ventilated
+
+        self.max_y_axis = max_y_axis
+
+        self.rates = tuple(
+            each.rate
+            for each in (hospitalized, icu, ventilated)
+        )
+        self.lengths_of_stay = tuple(
+            each.length_of_stay
+            for each in (hospitalized, icu, ventilated)
+        )
+
+        # Note: this should not be an integer.
+        # We're appoximating infected from what we do know.
+        # TODO market_share > 0, hosp_rate > 0
+        self.infected = infected = \
+            current_hospitalized / market_share / hospitalized.rate
+
+        self.detection_probability = \
+            known_infected / infected if infected > 1.e-7 else None
+
+        # TODO missing initial recovered value
+        self.recovered = 0.0
+
+        self.intrinsic_growth_rate = intrinsic_growth_rate = \
+            2.0 ** (1.0 / doubling_time) - 1.0 if doubling_time > 0.0 else 0.0
+
+        # TODO make this configurable, or more nuanced
+        self.recovery_days = recovery_days = 14.0
+
+        self.gamma = gamma = 1.0 / recovery_days
+
+        # Contact rate, beta
+        self.beta = beta = (
+            (intrinsic_growth_rate + gamma) /
+            susceptible * (1.0 - relative_contact_rate)
+        )  # {rate based on doubling time} / {initial susceptible}
+
+        # r_t is r_0 after distancing
+        self.r_t = beta / gamma * susceptible
+
+        # Simplify equation to avoid division by zero:
+        # self.r_naught = r_t / (1.0 - relative_contact_rate)
+        self.r_naught = (intrinsic_growth_rate + gamma) / gamma
+
+        # doubling time after distancing
+        # TODO constrain values np.log2(...) > 0.0
+        self.doubling_time_t = 1.0 / log2(beta * susceptible - gamma + 1)
+
+        self.dispositions = None
+        self.susceptible_v = self.infected_v = self.recovered_v = None
+        self.hospitalized_v = self.icu_v = self.ventilated_v = None
+
+        if n_days is not None:
+            self.n_days = n_days
+
+    @property
+    def n_days(self):
+        return self._n_days
+
+    @n_days.setter
+    def n_days(self, n_days: int):
+        self._n_days = n_days
+
+        s_v, i_v, r_v = sim_sir(
+            self.susceptible,
+            self.infected,
+            self.recovered,
+            self.beta,
+            self.gamma,
+            n_days,
+        )
+        self.susceptible_v, self.infected_v, self.recovered_v = s_v, i_v, r_v
+
+        self.dispositions = hospitalized_v, icu_v, ventilated_v = \
+            get_dispositions(i_v, self.rates, self.market_share)
+        self.hospitalized_v, self.icu_v, self.ventilated_v = \
+            hospitalized_v, icu_v, ventilated_v

src/penn_chime/presentation.py

@@ -4,8 +4,10 @@
 
 from .defaults import Constants, RateLos
 from .utils import add_date_column
-from .models import Parameters
-DATE_FORMAT = "%b, %d" # see https://strftime.org
+from .parameters import Parameters
+
+
+DATE_FORMAT = "%b, %d"  # see https://strftime.org
 
 
 hide_menu_style = """

src/penn_chime/utils.py

@@ -1,11 +1,25 @@
-from typing import Optional
+"""Utils."""
+
+from collections import namedtuple
 from datetime import datetime, timedelta
+from typing import Optional
+
 import numpy as np
 import pandas as pd
-import streamlit as st
 
-@st.cache
-def build_admissions_df(n_days, hosp, icu, vent) -> pd.DataFrame:
+
+# (0.02, 7) is 2%, 7 days
+# be sure to multiply by 100 when using as a default to the pct widgets!
+RateLos = namedtuple('RateLos', ('rate', 'length_of_stay'))
+
+
+def build_admissions_df(
+    n_days,
+    hosp,
+    icu,
+    vent,
+) -> pd.DataFrame:
+    """Build admis dataframe."""
     days = np.array(range(0, n_days + 1))
     data_dict = dict(zip(["day", "hosp", "icu", "vent"], [days, hosp, icu, vent]))
     projection = pd.DataFrame.from_dict(data_dict)
@@ -15,8 +29,13 @@ def build_admissions_df(n_days, hosp, icu, vent) -> pd.DataFrame:
     projection_admits["day"] = range(projection_admits.shape[0])
     return projection_admits
 
-@st.cache
-def build_census_df(projection_admits, hosp_los, icu_los, vent_los) -> pd.DataFrame:
+
+def build_census_df(
+    projection_admits,
+    hosp_los,
+    icu_los,
+    vent_los,
+) -> pd.DataFrame:
     """ALOS for each category of COVID-19 case (total guesses)"""
     n_days = np.shape(projection_admits)[0]
     los_dict = {
@@ -39,6 +58,7 @@ def build_census_df(projection_admits, hosp_los, icu_los, vent_los) -> pd.DataFr
     census_df = census_df.head(n_days)
     return census_df
 
+
 def add_date_column(
     df: pd.DataFrame,
     drop_day_column: bool = False,

src/test_app.py

@@ -1,3 +1,5 @@
+"""Tests."""
+
 import pytest
 import pandas as pd