Latest release

.github/workflows/build-deploy.yml

@@ -55,7 +55,8 @@ jobs:
           --rsv-hub-path $GITHUB_WORKSPACE/rsv-forecast-hub \
           --covid-hub-path $GITHUB_WORKSPACE/covid19-forecast-hub \
           --flu-metrocast-hub-path $GITHUB_WORKSPACE/flu-metrocast \
-          --NHSN
+          --NHSN \
+          --NSSP
 
     - name: Upload all processed data as a single artifact
       uses: actions/upload-artifact@v4

.github/workflows/parity-test.yml

@@ -27,7 +27,7 @@ jobs:
       - name: Install Python dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install pandas jsonschema
+          pip install pandas jsonschema requests
 
       - name: Set up R
         uses: r-lib/actions/setup-r@v2

.github/workflows/python-tests.yml

@@ -22,7 +22,7 @@ jobs:
       - name: Install dependencies
         run: |
           python -m pip install --upgrade pip
-          pip install pandas pytest jsonschema
+          pip install pandas pytest jsonschema requests
 
       - name: Run processor unit tests
         run: python -m pytest tests/test_processors.py

.github/workflows/scoringutils-wis-parity.yml

@@ -0,0 +1,39 @@
+name: Scoringutils WIS Parity
+
+on:
+  pull_request:
+  push:
+    branches:
+      - main
+
+jobs:
+  scoringutils-wis-parity:
+    runs-on: ubuntu-latest
+    defaults:
+      run:
+        working-directory: app
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Setup Node.js
+        uses: actions/setup-node@v4
+        with:
+          node-version: 'lts/*'
+          cache: 'npm'
+          cache-dependency-path: app/package-lock.json
+
+      - name: Set up R
+        uses: r-lib/actions/setup-r@v2
+        with:
+          use-public-rspm: true
+
+      - name: Install R packages
+        run: Rscript -e 'install.packages("scoringutils", repos = "https://cloud.r-project.org")'
+
+      - name: Install app dependencies
+        run: npm ci
+
+      - name: Run scoringutils WIS parity test
+        run: npm run test:scoringutils

README.md

@@ -2,15 +2,15 @@
 Authors: **Emily Przykucki**, Joseph Lemaitre, and others within ACCIDDA, the Atlantic Coast Center for Infectious Disease Dynamics and Analytics.
 
 * **stable version** https://www.RespiLens.com
-* **new features are developped on** https://staging.RespiLens.com, [GitHub](https://github.com/ACCIDDA/RespiLens-staging)
+* **new features are developed on** https://staging.RespiLens.com, [GitHub](https://github.com/ACCIDDA/RespiLens-staging)
 
 RespiLens is a responsive web app to visualize respiratory disease forecasts in the US, focused on accessibility for state health departments and the general public. The RSV, COVID-19 and flu views pull from CDC forecast hubs (collectively known as the Hubverse: [rsv-forecast-hub](https://github.com/CDCgov/rsv-forecast-hub), [covid19-forecast-hub](https://github.com/CDCgov/covid19-forecast-hub), and [FluSight-forecast-hub](https://github.com/cdcepi/FluSight-forecast-hub)), and consolidate pathogen data by location and date into one user-friendly plot. While other visualization tools for FluSight/RSV/COVID-19 Hubverse data may exist, many of them are geared towards academics instead of state health departments and the public, making them less accessible to people who need them. Our goal is to make a dashboard with these users in mind. Presently, RespiLens offers these features:
 - **Ability to link a certain view to a URL to share a forecast.** Select a pathogen, a location, and date(s), and then click the "Share View" button to copy the URL with your settings.
 - Site rebuilt daily, with new data Hub produced weekly during Hub forecasting seasons (year-round for COVID-19)
 - Ability to choose any number of dates to visualize on your plot
 - Ability to choose any number of contributing models to visualize on your plot
 - A view with National Healthcare Safety Network data, plotting almost 300 data categories (e.g., Number of Adult COVID-19 Admissions, or Percent Inpatient Beds Occupied by RSV Patients) 
-- **Forecastle**, a daily disease forecasting game, where you can attempt to predict hospitalizations counts for a specific senario and then get scored against participating models.
+- **Forecastle**, a daily disease forecasting game, where you can attempt to predict hospitalizations counts for a specific scenario and then get scored against participating models.
 - **MyRespiLens**, where you can visualize your own data for a specific location without the file leaving your machine.
 
 

app/.env

@@ -1,4 +1,6 @@
 # Tournament API Configuration
 # Replace with your deployed Google Apps Script Web App URL
 # See TOURNAMENT_SETUP.md for deployment instructions
-VITE_TOURNAMENT_API_URL=https://script.google.com/macros/s/AKfycbwB7LnE8DSk9S7ACLs20j65iB-9ryCXAiih2FlMwpeWDDE4pLZ1zF3RQilfrm6_byLU7w/exec
+VITE_EMILY_TOURNAMENT_API_URL=https://script.google.com/macros/s/AKfycby_pE9-KoA_bWjv9xIzNC1DF8jIrMPbQJ3I9P62RafivdQaHujnX2539tYFZtrn-nGRpw/exec
+VITE_EPIDEMICS10_TOURNAMENT_API_URL=https://script.google.com/macros/s/AKfycbwB7LnE8DSk9S7ACLs20j65iB-9ryCXAiih2FlMwpeWDDE4pLZ1zF3RQilfrm6_byLU7w/exec
+VITE_ALLHANDS_CHALLENGE_API_URL=https://script.google.com/macros/s/AKfycbzGBcCDAnAgnbCcyw0zSHSTNjlfgdiO5HyNCYn5gNGxhAjypgUcoufUTx2E7X1IOXyb/exec

app/package.json

@@ -8,6 +8,7 @@
     "build": "vite build",
     "build:staging": "vite build --mode staging",
     "lint": "eslint .",
+    "test:scoringutils": "node scripts/test-scoringutils-wis.mjs",
     "preview": "vite preview",
     "format": "prettier . --write",
     "format:check": "prettier . --check --log-level warn",

app/scripts/test-scoringutils-wis.mjs

@@ -0,0 +1,158 @@
+import assert from "node:assert/strict";
+import { spawnSync } from "node:child_process";
+
+import { calculateWIS } from "../src/utils/forecastleScoring.js";
+import { calculateWIS as calculateSharedWIS } from "../src/lib/forecast-components/scoring.js";
+
+const rscriptCandidates = [process.env.SCORINGUTILS_RSCRIPT, "Rscript"].filter(
+  Boolean,
+);
+
+const findScoringutilsRscript = () => {
+  for (const rscript of rscriptCandidates) {
+    const check = spawnSync(
+      rscript,
+      [
+        "-e",
+        "quit(status = !requireNamespace('scoringutils', quietly = TRUE))",
+      ],
+      { encoding: "utf8" },
+    );
+    if (check.status === 0) {
+      return rscript;
+    }
+  }
+  return null;
+};
+
+const rscript = findScoringutilsRscript();
+
+if (!rscript) {
+  if (process.env.CI) {
+    throw new Error("No Rscript with scoringutils found.");
+  } else {
+    console.log(
+      "Skipping scoringutils WIS parity test: no Rscript with scoringutils found.",
+    );
+    process.exit(0);
+  }
+}
+
+const cases = [
+  {
+    observed: 100,
+    median: 105,
+    lower50: 90,
+    upper50: 115,
+    lower95: 80,
+    upper95: 140,
+  },
+  {
+    observed: 70,
+    median: 105,
+    lower50: 90,
+    upper50: 115,
+    lower95: 80,
+    upper95: 140,
+  },
+  {
+    observed: 160,
+    median: 105,
+    lower50: 90,
+    upper50: 115,
+    lower95: 80,
+    upper95: 140,
+  },
+];
+
+const rProgram = `
+library(scoringutils)
+cases <- list(
+  list(observed=100, predicted=c(80, 90, 105, 115, 140)),
+  list(observed=70, predicted=c(80, 90, 105, 115, 140)),
+  list(observed=160, predicted=c(80, 90, 105, 115, 140))
+)
+quantile_level <- c(0.025, 0.25, 0.5, 0.75, 0.975)
+for (case in cases) {
+  result <- wis(
+    observed = case$observed,
+    predicted = case$predicted,
+    quantile_level = quantile_level,
+    separate_results = TRUE
+  )
+  cat(sprintf(
+    "%.12f,%.12f,%.12f,%.12f\\n",
+    result$wis,
+    result$dispersion,
+    result$underprediction,
+    result$overprediction
+  ))
+}
+`;
+
+const result = spawnSync(rscript, ["-e", rProgram], { encoding: "utf8" });
+
+if (result.status !== 0) {
+  throw new Error(
+    `scoringutils parity reference failed:\n${result.stderr || result.stdout}`,
+  );
+}
+
+const expected = result.stdout
+  .trim()
+  .split("\n")
+  .map((line) => {
+    const [wis, dispersion, underprediction, overprediction] = line
+      .split(",")
+      .map(Number);
+    return { wis, dispersion, underprediction, overprediction };
+  });
+
+const assertClose = (actual, expectedValue, label) => {
+  assert.ok(
+    Math.abs(actual - expectedValue) < 1e-9,
+    `${label}: expected ${expectedValue}, got ${actual}`,
+  );
+};
+
+cases.forEach((testCase, index) => {
+  const forecastle = calculateWIS(
+    testCase.observed,
+    testCase.median,
+    testCase.lower50,
+    testCase.upper50,
+    testCase.lower95,
+    testCase.upper95,
+  );
+  const shared = calculateSharedWIS(
+    testCase.observed,
+    testCase.median,
+    testCase.lower50,
+    testCase.upper50,
+    testCase.lower95,
+    testCase.upper95,
+  );
+
+  for (const implementation of [forecastle, shared]) {
+    assertClose(implementation.wis, expected[index].wis, `case ${index} wis`);
+    assertClose(
+      implementation.dispersion,
+      expected[index].dispersion,
+      `case ${index} dispersion`,
+    );
+    assertClose(
+      implementation.underprediction,
+      expected[index].underprediction,
+      `case ${index} underprediction`,
+    );
+    assertClose(
+      implementation.overprediction,
+      expected[index].overprediction,
+      `case ${index} overprediction`,
+    );
+  }
+});
+
+assert.equal(calculateWIS(100, NaN, 90, 115, 80, 140), null);
+
+console.log(`scoringutils WIS parity passed using ${rscript}`);

app/src/components/DataVisualizationContainer.jsx

@@ -493,6 +493,44 @@ const DataVisualizationContainer = ({ disableSeo = false }) => {
         </>
       ),
     },
+    nsspall: {
+      title: (
+        <Group gap="sm">
+          <Title order={4}>
+            National Syndromic Surveillance Program (NSSP)
+          </Title>
+        </Group>
+      ),
+      buttonLabel: "About NSSP Data",
+      content: (
+        <>
+          <p>
+            Data for the RespiLens NSSP view comes from the CDC's{" "}
+            <a
+              href="https://www.cdc.gov/nssp/index.html"
+              target="_blank"
+              rel="noopener noreferrer"
+            >
+              National Syndromic Surveillance Program,
+            </a>{" "}
+            which integrates data from many sources to help consolidate
+            complete, accurate, and timely health data. Syndromic surveillance
+            is a collaborative effort among local and state health departments,
+            CDC, and partners.
+          </p>
+          <div>
+            <Title order={4} mb="xs">
+              Data
+            </Title>
+            <p>
+              RespiLens' display of NSSP data shows percent of hospital visits
+              attributable to RSV, COVID-19, and influenza across 50 states +
+              Washington D.C.; where 25 states have county-level visualization.
+            </p>
+          </div>
+        </>
+      ),
+    },
   };
 
   const currentAboutConfig = aboutHubConfig[viewType];

app/src/components/Documentation.jsx

@@ -332,7 +332,7 @@ const Documentation = () => {
               successfully build a RespiLens projections JSON file. The
               conversion pipeline's shortlist of requirements is below. Please
               note that if you pulled your data directly from a Hubverse hub{" "}
-              <code>.csv</code> file, it likley already conforms to conversion
+              <code>.csv</code> file, it likely already conforms to conversion
               requirements and you can skip to the next step!
             </Text>
             <Text>