Merge branch 'develop' into issue_155

Author: mdbecker

.github/ISSUE_TEMPLATE/general-bug-report.md

@@ -0,0 +1,27 @@
+---
+name: General bug report
+about: Create an report to help us improve.
+title: [bug]
+labels: 'bug'
+assignees: ''
+
+---
+
+<!--
+Before you begin, please help us manage volume by checking if this could be submitted another way:
+- Usage questions? Ask in Slack [#chime-help](https://codeforphilly.org/chat?channel=chime-help).
+- Feature ideas? Propose in Slack for discussion.
+- Is this something you can debug and fix? Pull requests are very welcome.
+-->
+
+### Description:
+
+### Steps to reproduce:
+1.
+2.
+3.
+
+### Expected behavior:
+
+### What I got instead:
+(screenshots if applicable)

.github/ISSUE_TEMPLATE/model.md

@@ -0,0 +1,26 @@
+---
+name: Bug report for model
+about: Feedback or bug report for the epidemiological model and analysis.
+title: [model]
+labels: 'models'
+assignees: ''
+
+---
+
+<!--
+Please note: Any changes to the model have a huge impact on rapidly evolving hospital system & public health decisions. The current model has been in use for a while now, and it has been validated against other similar models, so any changes to the model must meet a very high bar.
+
+However, these 2 types of issue reports are very welcome:
+- Bugs causing this model to produce invalid results. In this case, please include details and a suggested fix.
+- If this model is producing a significantly different result than another well-known epidemiological model. In this case, please include proof of this difference and a suggested fix to our approach.
+
+For questions or early discussion, please join us in [#chime-analysis](https://codeforphilly.org/chat?channel=chime-analysis) in Slack instead.
+-->
+
+### Summary
+
+
+### Additional details
+
+
+### Suggested fix

README.md

@@ -7,7 +7,7 @@ The **C**OVID-19 **H**ospital **I**mpact **M**odel for **E**pidemics ([penn-chim
 ## Background
 The [CHIME](https://penn-chime.phl.io/) (COVID-19 Hospital Impact Model for Epidemics) Application is designed to assist hospitals and public health officials with understanding hospital capacity needs as they relate to the COVID pandemic. CHIME enables capacity planning by providing estimates of total daily (i.e. new) and running totals of (i.e. census) inpatient hospitalizations, ICU admissions, and patients requiring ventilation. These estimates are generated using a [SIR (Susceptible, Infected, Recovered)](https://mathworld.wolfram.com/SIRModel.html) model, a standard epidemiological modeling technique. Our model has been validated by several epidemiologists including [Michael Z. Levy, PhD](https://www.dbei.med.upenn.edu/bio/michael-z-levy-phd), Associate Professor of Epidemiology, Department of Biostatistics, Epidemiology and Informatics at the Perelman School of Medicine.
 
-Originally developed in `github.com/pennsignals/chime`, active development is now at `github.com/pennsignals/chime`.
+Originally developed in `github.com/pennsignals/chime`, active development is now at `github.com/CodeForPhilly/chime`.
 
 ### Documentation
 

docs/getting-started/try-online.md

@@ -12,7 +12,7 @@ Try this link: https://pennchime.herokuapp.com/
 
 ### Run Your Own Copy Locally
 
-If you're comfortable working on your computer's command line, you can head over to the [Getting Started: Run Locally](getting-started/run-locally.md) guide to run a private instance on your own computer.
+If you're comfortable working on your computer's command line, you can head over to the [Getting Started: Run Locally](run-locally.md) guide to run a private instance on your own computer.
 
 ### Run a Shared Copy Online for Your Organization
 

script/server

@@ -1,4 +1,4 @@
-#!/usr/bin/env sh
+#!/usr/bin/env bash
 
 # script/server: Launch the application and any extra required processes
 #                locally.

src/app.py

@@ -17,11 +17,13 @@
     write_footer,
 )
 from penn_chime.settings import DEFAULTS
-from penn_chime.models import sim_sir_df, build_admissions_df, build_census_df
-from penn_chime.charts import (additional_projections_chart,
-                               admitted_patients_chart,
-                               new_admissions_chart,
-                               chart_descriptions)
+from penn_chime.models import SimSirModel
+from penn_chime.charts import (
+    additional_projections_chart,
+    admitted_patients_chart,
+    new_admissions_chart,
+    chart_descriptions
+)
 
 # This is somewhat dangerous:
 # Hide the main menu with "Rerun", "run on Save", "clear cache", and "record a screencast"
@@ -30,55 +32,52 @@
 st.markdown(hide_menu_style, unsafe_allow_html=True)
 
 p = display_sidebar(st, DEFAULTS)
+m = SimSirModel(p)
 
-display_header(st, p)
+display_header(st, m, p)
 
 if st.checkbox("Show more info about this tool"):
     notes = "The total size of the susceptible population will be the entire catchment area for Penn Medicine entities (HUP, PAH, PMC, CCH)"
-    show_more_info_about_this_tool(st=st, parameters=p, inputs=DEFAULTS, notes=notes)
-
-
-# begin format data
-admissions_df = build_admissions_df(p=p)  # p.n_days, *p.dispositions)
-census_df = build_census_df(admissions_df, parameters=p)
-# end format data
+    show_more_info_about_this_tool(st=st, model=m, parameters=p, defaults=DEFAULTS, notes=notes)
 
 st.subheader("New Admissions")
 st.markdown("Projected number of **daily** COVID-19 admissions at Penn hospitals")
-new_admit_chart = new_admissions_chart(alt, admissions_df, parameters=p)
+new_admit_chart = new_admissions_chart(alt, m.admits_df, parameters=p)
 st.altair_chart(
-    new_admit_chart, use_container_width=True
+    new_admissions_chart(alt, m.admits_df, parameters=p),
+    use_container_width=True,
 )
 
-st.markdown(chart_descriptions(new_admit_chart))
+st.markdown(chart_descriptions(new_admit_chart, p.labels))
 
 if st.checkbox("Show Projected Admissions in tabular form"):
     if st.checkbox("Show Daily Counts"):
-        draw_projected_admissions_table(st, admissions_df, as_date=p.as_date, daily_count=True)
+        draw_projected_admissions_table(st, m.admits_df, p.labels, as_date=p.as_date, daily_count=True)
     else:
-        draw_projected_admissions_table(st, admissions_df, as_date=p.as_date, daily_count=False)
+        draw_projected_admissions_table(st, m.admits_df, p.labels, as_date=p.as_date, daily_count=False)
     build_download_link(st,
         filename="projected_admissions.csv",
-        df=admissions_df,
+        df=m.admits_df,
         parameters=p
     )
 st.subheader("Admitted Patients (Census)")
 st.markdown(
     "Projected **census** of COVID-19 patients, accounting for arrivals and discharges at Penn hospitals"
 )
-census_chart = admitted_patients_chart(alt=alt, census=census_df, parameters=p)
+census_chart = admitted_patients_chart(alt=alt, census=m.census_df, parameters=p)
 st.altair_chart(
-    census_chart, use_container_width=True
+    admitted_patients_chart(alt=alt, census=m.census_df, parameters=p),
+    use_container_width=True,
 )
-st.markdown(chart_descriptions(census_chart, suffix=" Census"))
+st.markdown(chart_descriptions(census_chart, p.labels, suffix=" Census"))
 if st.checkbox("Show Projected Census in tabular form"):
     if st.checkbox("Show Daily Census Counts"):
-        draw_census_table(st, admissions_df, as_date=p.as_date, daily_count=True)
+        draw_census_table(st, m.census_df, p.labels, as_date=p.as_date, daily_count=True)
     else:
-        draw_census_table(st, census_df, as_date=p.as_date, daily_count=False)
+        draw_census_table(st, m.census_df, p.labels, as_date=p.as_date, daily_count=False)
     build_download_link(st,
         filename="projected_census.csv",
-        df=census_df,
+        df=m.census_df,
         parameters=p
     )
 
@@ -87,9 +86,9 @@
 )
 if st.checkbox("Show Additional Projections"):
     show_additional_projections(
-        st, alt, additional_projections_chart, parameters=p
+        st, alt, additional_projections_chart, model=m, parameters=p
     )
     if st.checkbox("Show Raw SIR Simulation Data"):
-        draw_raw_sir_simulation_table(st, parameters=p)
+        draw_raw_sir_simulation_table(st, model=m, parameters=p)
 write_definitions(st)
 write_footer(st)

src/cli.py

@@ -9,8 +9,8 @@
 from pandas import DataFrame
 
 from penn_chime.parameters import Parameters
+from penn_chime.models import SimSirModel
 from penn_chime.utils import RateLos
-from penn_chime.models import build_admissions_df, build_census_df
 
 class FromFile(Action):
     """From File."""
@@ -108,30 +108,22 @@ def main():
         doubling_time=a.doubling_time,
         known_infected=a.known_infected,
         market_share=a.market_share,
+        n_days=a.n_days,
         relative_contact_rate=a.relative_contact_rate,
         susceptible=a.susceptible,
-        n_days=a.n_days,
+
         hospitalized=RateLos(a.hospitalized_rate, a.hospitalized_los),
         icu=RateLos(a.icu_rate, a.icu_los),
         ventilated=RateLos(a.ventilated_rate, a.ventilated_los),
     )
 
-    raw_df = DataFrame(
-        {
-            "Susceptible": p.susceptible_v,
-            "Infected": p.infected_v,
-            "Recovered": p.recovered_v,
-        }
-    )
-
-    admits_df = build_admissions_df(p)
-    census_df = build_census_df(admits_df, p)
+    m = SimSirModel(p)
 
     prefix = a.prefix
     for df, name in (
-        (raw_df, "raw"),
-        (admits_df, "admits"),
-        (census_df, "census"),
+        (m.raw_df, "raw"),
+        (m.admits_df, "admits"),
+        (m.census_df, "census"),
     ):
         df.to_csv(prefix + name + ".csv")
 

src/penn_chime/charts.py

@@ -1,12 +1,13 @@
 
 from math import ceil
 import datetime
+
 from altair import Chart  # type: ignore
 import pandas as pd  # type: ignore
-import numpy as np  # type: ignore
 
 from .parameters import Parameters
 from .utils import add_date_column
+from .presentation import DATE_FORMAT
 
 
 def new_admissions_chart(
@@ -26,13 +27,14 @@ def new_admissions_chart(
     tooltip_dict = {False: "day", True: "date:T"}
     if as_date:
         projection_admits = add_date_column(projection_admits)
-        x_kwargs = {"shorthand": "date:T", "title": "Date"}
+        x_kwargs = {"shorthand": "date:T", "title": "Date", "axis": alt.Axis(format=(DATE_FORMAT))}
     else:
         x_kwargs = {"shorthand": "day", "title": "Days from today"}
 
+    # TODO fix the fold to allow any number of dispositions
     return (
         alt.Chart(projection_admits.head(plot_projection_days))
-        .transform_fold(fold=["Hospitalized", "ICU", "Ventilated"])
+        .transform_fold(fold=["hospitalized", "icu", "ventilated"])
         .mark_line(point=True)
         .encode(
             x=alt.X(**x_kwargs),
@@ -58,7 +60,7 @@ def admitted_patients_chart(
     as_date = parameters.as_date
     if as_date:
         census = add_date_column(census)
-        x_kwargs = {"shorthand": "date:T", "title": "Date"}
+        x_kwargs = {"shorthand": "date:T", "title": "Date", "axis": alt.Axis(format=(DATE_FORMAT))}
         idx = "date:T"
     else:
         x_kwargs = {"shorthand": "day", "title": "Days from today"}
@@ -70,9 +72,10 @@ def admitted_patients_chart(
         y_scale.domain = (0, max_y_axis)
         y_scale.clamp = True
 
+    # TODO fix the fold to allow any number of dispositions
     return (
         alt.Chart(census.head(plot_projection_days))
-        .transform_fold(fold=["Hospitalized", "ICU", "Ventilated"])
+        .transform_fold(fold=["hospitalized", "icu", "ventilated"])
         .mark_line(point=True)
         .encode(
             x=alt.X(**x_kwargs),
@@ -89,19 +92,23 @@ def admitted_patients_chart(
 
 
 def additional_projections_chart(
-    alt, parameters: Parameters
+    alt, model, parameters
 ) -> Chart:
-    i = parameters.infected_v
-    r = parameters.recovered_v
-    dat = pd.DataFrame({"Infected": i, "Recovered": r})
+
+    # TODO use subselect of df_raw instead of creating a new df
+    raw_df = model.raw_df
+    dat = pd.DataFrame({
+        "infected": raw_df.infected,
+        "recovered": raw_df.recovered
+    })
     dat["day"] = dat.index
 
     as_date = parameters.as_date
     max_y_axis = parameters.max_y_axis
 
     if as_date:
         dat = add_date_column(dat)
-        x_kwargs = {"shorthand": "date:T", "title": "Date"}
+        x_kwargs = {"shorthand": "date:T", "title": "Date", "axis": alt.Axis(format=(DATE_FORMAT))}
     else:
         x_kwargs = {"shorthand": "day", "title": "Days from today"}
 
@@ -113,7 +120,7 @@ def additional_projections_chart(
 
     return (
         alt.Chart(dat)
-        .transform_fold(fold=["Infected", "Recovered"])
+        .transform_fold(fold=["infected", "recovered"])
         .mark_line()
         .encode(
             x=alt.X(**x_kwargs),
@@ -125,7 +132,7 @@ def additional_projections_chart(
     )
 
 
-def chart_descriptions(chart: Chart, suffix: str = ""):
+def chart_descriptions(chart: Chart, labels, suffix: str = ""):
     """
 
     :param chart: Chart: The alt chart to be used in finding max points
@@ -136,7 +143,7 @@ def chart_descriptions(chart: Chart, suffix: str = ""):
     """
     messages = []
 
-    cols = ["Hospitalized", "ICU", "Ventilated"]
+    cols = ["hospitalized", "icu", "ventilated"]
     asterisk = False
     day = "date" if "date" in chart.data.columns else "day"
 
@@ -152,7 +159,7 @@ def chart_descriptions(chart: Chart, suffix: str = ""):
 
         messages.append(
             "{}{} peaks at {:,} on day {}{}".format(
-                col,
+                labels[col],
                 suffix,
                 ceil(chart.data[col].max()),
                 on,

src/penn_chime/defaults.py

@@ -25,25 +25,34 @@ def __init__(
         current_hospitalized: int,
         doubling_time: int,
         known_infected: int,
-        n_days: int,
         relative_contact_rate: int,
         region: Regions,
+
         hospitalized: RateLos,
         icu: RateLos,
         ventilated: RateLos,
+
+        as_date: bool = False,
         market_share: float = 1.0,
+        max_y_axis: int = None,
+        n_days: int = 60,
+        recovery_days: int = 14,
     ):
         self.region = region
-        self.known_infected = known_infected
         self.current_hospitalized = current_hospitalized
+        self.known_infected = known_infected
         self.doubling_time = doubling_time
-        self.market_share = market_share
         self.relative_contact_rate = relative_contact_rate
 
         self.hospitalized = hospitalized
         self.icu = icu
         self.ventilated = ventilated
+
+        self.as_date = as_date
+        self.market_share = market_share
+        self.max_y_axis = max_y_axis
         self.n_days = n_days
+        self.recovery_days = recovery_days
 
     def __repr__(self) -> str:
         return f"Constants(susceptible_default: {self.region.susceptible}, known_infected: {self.known_infected})"