SMHS Epidemiology - Revision history

Dinov: /* Realistic Epidemiological Quantification */

2026-02-17T19:49:21Z

‎Realistic Epidemiological Quantification

2026-02-12T14:49:21Z

‎References

2026-02-10T20:56:14Z

‎References

2026-02-10T20:52:10Z

‎Problem 7

2026-02-10T20:36:40Z

‎Problem 6: Diagnostic Test Accuracy via ROC Analysis

2026-02-10T20:34:54Z

‎Problem 5: Assessing Confounding in Observational Data

2026-02-10T20:33:34Z

‎Problem 5: Assessing Confounding in Observational Data

2026-02-10T20:21:24Z

‎Realistic Epidemiological Quantification

2026-02-10T20:13:33Z

‎Odds Ratio (OR)

2026-02-10T20:12:37Z

‎Additional Considerations

@@ Line 803: / Line 803: @@
 sim_data <- bind_rows(history)
-# --- Visualization: Spatial Intensity (Fixed) ---
+# --- Visualization: Spatial Intensity ---
 ggplot(sim_data %>% filter(state == "I"), aes(x = x, y = y)) +
-     stat_density_2d(aes(fill = after_stat(level)),  # Updated syntax here
+  geom_density_2d_filled(
-                    geom = "polygon",
+     aes(fill = after_stat(level)),
-                    alpha = 0.3) +
+    alpha = 0.6,
-    geom_point(aes(color = genome), alpha = 0.6) +
+    contour_var = "ndensity"   # optional: normalize for better contrast
-    scale_fill_viridis_c() +
+  ) +
-    theme_minimal() +
+  geom_point(aes(color = genome), alpha = 0.9) +
-    labs(title = "Viral Hotspots and Genomic Clustering") +
+  scale_fill_viridis_d() +      # discrete fill for the bands
-    theme(legend.position = "none")  # Hide genome legend if too many unique strings
+  theme_minimal() +
 </pre>

@@ Line 1,608: / Line 1,608: @@
 === References ===
 # Rothman KJ, Greenland S, Lash TL. Modern Epidemiology. 3rd ed. Lippincott, 2008.
 # Clayton D, Hills M. Statistical Models in Epidemiology. Oxford, 2013.

@@ Line 1,615: / Line 1,615: @@
 <hr>
-* SOCR Home page: http://www.socr.umich.edu
+* SOCR Home page: https://socr.umich.edu
-{{translate|pageName=http://wiki.socr.umich.edu/index.php?title=SMHS_Epidemiology}}
+{{translate|pageName=https://wiki.socr.umich.edu/index.php?title=SMHS_Epidemiology}}

@@ Line 1,231: / Line 1,231: @@
 for (t in 1:timesteps) {
-   # FIX 2: Assignment operator (<-) to update movement
+   # Assignment operator (<-) to update movement
    pop$x <- pop$x + rnorm(n_agents, 0, 2)
    pop$y <- pop$y + rnorm(n_agents, 0, 2)
-−
    # Optional: Keep agents within map boundaries
    pop$x <- pmax(0, pmin(map_size, pop$x))

@@ Line 1,140: / Line 1,140: @@
 ==== Problem 6: Diagnostic Test Accuracy via ROC Analysis ====
-'''Scenario''': A continuous biomarker is measured in 100 diseased and 100 non-diseased individuals. Determine the optimal cutoff, sensitivity, and specificity using Youden’s index.
+'''Scenario''': A continuous biomarker is measured in 100 diseased and 100 non-diseased individuals. Determine the optimal cutoff, sensitivity, and specificity using [https://doi.org/10.1002/bimj.200710415 Youden’s index].
 <pre>
@@ Line 1,166: / Line 1,166: @@
    cat("Specificity:", round(opt[["specificity"]], 3), "\n")
 </pre>
 ====Problem 7====

@@ Line 1,136: / Line 1,136: @@
    if (abs(confounding_pct) > 10) {
     cat("→ Substantial confounding detected (change > 10%).\n")
+   } else cat("→ No substantial confounding detected (change < 10%).\n")
 </pre>

@@ Line 1,106: / Line 1,106: @@
 '''Scenario''': Evaluate whether age and sex confound the association between an exposure and disease outcome.
    set.seed(123)
    n <- 5000
@@ Line 1,134: / Line 1,137: @@
     cat("→ Substantial confounding detected (change > 10%).\n")
+   }
 ==== Problem 6: Diagnostic Test Accuracy via ROC Analysis ====

@@ Line 684: / Line 684: @@
 <pre>
 # Example quantification of lineage prevalence
-lineage_summary <- sim_data %>%
+library(dplyr)
-  filter(state == "I") %>%
+library(ggplot2)
-  group_material(time, genome) %>%
-  summarise(count = n()) %>%
+lineage_summary_plot <- sim_data %>%
-  ggplot(aes(x = time, y = count, fill = as.factor(genome))) +
+    filter(state == "I") %>%
-  geom_area() +
+    # Use group_by to organize the data by time and genome
-  labs(title = "Genomic Lineage Succession", x = "Days", y = "Active Cases")
+    group_by(time, genome) %>%
-lineage_summary
+    summarise(count = n(), .groups = "drop") %>%
 </pre>

@@ Line 387: / Line 387: @@
-Formula:
+Formula: <math>OR = \frac{ad}{bc}</math>
 OR = \frac{ad}{bc}
-(Odds Ratio = [Exposed Cases × Unexposed Non-cases] / [Exposed Non-cases × Unexposed Cases])
+: Odds Ratio = [Exposed Cases × Unexposed Non-cases] / [Exposed Non-cases × Unexposed Cases].
 * '''Definition''': The ratio of the odds of an outcome in the exposed group to the odds in the unexposed group. OR approximates RR when the outcome is rare.

← Older revision		Revision as of 20:12, 10 February 2026
Line 463:		Line 463:


−	The following code uses a custom function to handle both benefit (reduction in risk) and harm (increase in risk) scenarios dynamically.	+	The following code uses a custom function to handle both ''benefit'' (reduction in risk) and ''harm'' (increase in risk) scenarios dynamically.

	<pre>		<pre>