fix comment start/end

Author: nmdefries

slides/day1-afternoon.qmd

@@ -1135,7 +1135,7 @@ snapshots <- map(versions, \(v) {
 <!--
 ## Backfill American edition
 
-* Again, we can see a similar systematic underestimation problem for COVID-19 mortality rates in CA. <!-- 2023-2024 -->
+* Again, we can see a similar systematic underestimation problem for COVID-19 mortality rates in CA.
 
 * This plot also illustrates the [**revision process**]{.primary} - how the reported mortality changes & increases across multiple updates until it stabilizes at the final value (black line).
 
@@ -2405,7 +2405,7 @@ carefully tomorrow.
 
 * Ex. using the number of reported COVID-19 cases in British Columbia between Jan. 2020 and Apr. 15, 2023.
 
-<!-- This data is the number of reported COVID-19 cases in British Columbia between January 2020 and April 15, 2023. The values are.up-to-date as of August 2023. -->
+This data is the number of reported COVID-19 cases in British Columbia between January 2020 and April 15, 2023. The values are.up-to-date as of August 2023.
 ```{r rtestim}
 #| fig-width: 9
 #| fig-height: 3

slides/day1-morning.qmd

@@ -414,7 +414,7 @@ lat_ca_june1 <- dv_ca |>
   pull() |> as.numeric()
 ```
 
-* On June 1, this signal is subject to [**revision**]{.primary} <!--over time (ex. consider Dec. 1's `percent_cli` across `version`):-->
+* On June 1, this signal is subject to [**revision**]{.primary} over time (ex. consider Dec. 1's `percent_cli` across `version`):
 
 ```{r revision-ca-ex}
 #| echo: false
@@ -601,13 +601,11 @@ In Bexar County, Texas, during the summer of 2020...
 * Auxilliary signals show no such dramatic increase
 * Reports themselves may not be trustworthy without context
 
-<!--
 
 * **Left**: On July 16, 4,810 [backlogged cases]{.primary} were reported, reflecting a 2-week delay. This caused a prolonged spike due to the 7-day trailing average applied to the counts.
 
 * **Right**: CTIS estimates of CLI-in-community showed more stable underlying trends.
 
--->
 
 ```{r fig4-Reinhart}
 #| echo: false
@@ -701,10 +699,10 @@ Added value
 ## Goals of Delphi Epidata platform and repository
 
 1. **Provide many aggregated epi-surveillance time-series ("epi-signals")**
-    + Mirror signals from other sources, especially if revisions are not tracked <!-- E.g. CDC's own NSSP, NWSS -->
+    + Mirror signals from other sources, especially if revisions are not tracked
     + Be the national historical repository of record & preserve the raw data
 
-<!-- Hence: include also signals available elsewhere, especially if they don't keep data revisions - E.g. CDC's own NSSP, NWSS -->
+Hence: include also signals available elsewhere, especially if they don't keep data revisions - E.g. CDC's own NSSP, NWSS
 
 2. **Be the go-to place for epi-signal discovery, including those held elsewhere**
 
@@ -715,6 +713,7 @@ Added value
 ::: {.callout-important appearance="simple"}
 Make epi-surveillance more nimble, complete, standardized, robust, and real-time
 :::
+
 -->
 
 
@@ -1340,7 +1339,7 @@ tail(dv_pa_issues_all) |>
 * Avoid queries with too-late minimum too-early maximum issue
 * Could be misleading results
 
-<!-- * We caution against starting queries with too late a minimum issue or too early maximum issue, as it could lead to incorrect or misleading results. You're safest bet to capture all issues is on the next slide... -->
+* We caution against starting queries with too late a minimum issue or too early maximum issue, as it could lead to incorrect or misleading results. You're safest bet to capture all issues is on the next slide...
 
 ## Obtaining all issues for one state
 
@@ -1420,14 +1419,15 @@ dv_state_default <- pub_covidcast(
 
 -->
 
+<!--
 ## Obtaining one issue for all states
-
+-->
 <!--
 [**Final question:**]{.primary}  What do you think happens when we adopt a "do nothing" approach to `geo_values` and `issues` (take both of them out)?
 
 <small> [**Hint**]{.primary}: remember a couple slides ago, when we removed `as_of`, we got the most recent estimate for PA. </small>
 -->
-
+<!--
 Defaults are intended to be "what you would expect"
 
 ```{r one-issue-all-states}
@@ -1508,7 +1508,7 @@ head(epidata) |> select(geo_value, signal, source, time_value, issue, lag, value
 ```
 </div>
 -->
--->
+
 
 
 ## Main takeaways

slides/day2-morning.qmd

@@ -237,7 +237,7 @@ $y_{new,t}$ over times $t = 1, \dots, N$.
 
   * mean absolute scaled error (MASE)
 
-<--
+<!--
 ## Error metrics: MSE and MAE
 
 $$MSE = \frac{1}{N} \sum_{t=1}^N (y_{new, t}- \hat y_{new, t})^2$$
@@ -2077,4 +2077,3 @@ with $\theta_0 = 1$
 Assumes an $ARMA(p, q)$ model on the
 $d^{th}$-order differences of $y_t$
 -->
--->