dan's simple suggestions

Author: dsweber2

NEWS.md

@@ -38,7 +38,7 @@ Pre-1.0.0 numbering scheme: 0.x will indicate releases, while 0.0.x will indicat
 - Replace `dist_quantiles()` with `hardhat::quantile_pred()`
 - Allow `quantile()` to threshold to an interval if desired (#434)
 - `arx_forecaster()` detects if there's enough data to predict
-- Add `plot_data` to `autoplot` so that forecasts can be plotted against the values they're predicting
+- Add `observed_response` to `autoplot` so that forecasts can be plotted against the values they're predicting
 
 ## Bug fixes
 

R/autoplot.R

@@ -16,7 +16,7 @@ ggplot2::autoplot
 #' @param object,x An `epi_workflow`
 #' @param predictions A data frame with predictions. If `NULL`, only the
 #'   original data is shown.
-#' @param plot_data An epi_df of the data to plot against. This is for the case
+#' @param observed_response An epi_df of the data to plot against. This is for the case
 #'   where you have the actual results to compare the forecast against.
 #' @param .levels A numeric vector of levels to plot for any prediction bands.
 #'   More than 3 levels begins to be difficult to see.
@@ -85,7 +85,7 @@ NULL
 autoplot.epi_workflow <- function(
     object,
     predictions = NULL,
-    plot_data = NULL,
+    observed_response = NULL,
     .levels = c(.5, .8, .9), ...,
     .color_by = c("all_keys", "geo_value", "other_keys", ".response", "all", "none"),
     .facet_by = c(".response", "other_keys", "all_keys", "geo_value", "all", "none"),
@@ -120,10 +120,10 @@ autoplot.epi_workflow <- function(
   } else {
     new_name_y <- names(y)
   }
-  if (is.null(plot_data)) {
+  if (is.null(observed_response)) {
     # the outcome has shifted, so we need to shift it forward (or back)
     # by the corresponding amount
-    plot_data <- bind_cols(mold$extras$roles[mold_roles %in% keys], y)
+    observed_response <- bind_cols(mold$extras$roles[mold_roles %in% keys], y)
     if (starts_with_impl("ahead_", names(y))) {
       shift <- as.numeric(old_name_y[2])
     } else if (starts_with_impl("lag_", names(y))) {
@@ -133,21 +133,22 @@ autoplot.epi_workflow <- function(
       new_name_y <- names(y)
       shift <- 0
     }
-    plot_data <- rename(plot_data, !!new_name_y := !!names(y))
+    observed_response <- rename(observed_response, !!new_name_y := !!names(y))
     if (!is.null(shift)) {
-      plot_data <- mutate(plot_data, time_value = time_value + shift)
+      observed_response <- mutate(observed_response, time_value = time_value + shift)
     }
     other_keys <- setdiff(key_colnames(object), c("geo_value", "time_value"))
-    plot_data <- as_epi_df(plot_data,
+    observed_response <- as_epi_df(observed_response,
       as_of = object$fit$meta$as_of,
       other_keys = other_keys
     )
   }
   if (is.null(predictions)) {
     return(autoplot(
-      plot_data, new_name_y,
+      observed_response, new_name_y,
       .color_by = .color_by, .facet_by = .facet_by, .base_color = .base_color,
-      .facet_filter = {{ .facet_filter }}
+      .facet_filter = {{ .facet_filter }},
+      .max_facets = .max_facets
     ))
   }
 
@@ -157,27 +158,29 @@ autoplot.epi_workflow <- function(
     }
     predictions <- rename(predictions, time_value = target_date)
   }
-  pred_cols_ok <- hardhat::check_column_names(predictions, key_colnames(plot_data))
+  pred_cols_ok <- hardhat::check_column_names(predictions, key_colnames(observed_response))
   if (!pred_cols_ok$ok) {
     cli_warn(c(
       "`predictions` is missing required variables: {.var {pred_cols_ok$missing_names}}.",
       i = "Plotting the original data."
     ))
     return(autoplot(
-      plot_data, !!new_name_y,
+      observed_response, !!new_name_y,
       .color_by = .color_by, .facet_by = .facet_by, .base_color = .base_color,
-      .facet_filter = {{ .facet_filter }}
+      .facet_filter = {{ .facet_filter }},
+      .max_facets = .max_facets
     ))
   }
 
   # First we plot the history, always faceted by everything
-  bp <- autoplot(plot_data, !!new_name_y,
+  bp <- autoplot(observed_response, !!new_name_y,
     .color_by = "none", .facet_by = "all_keys",
-    .base_color = "black", .facet_filter =  {{ .facet_filter }}
+    .base_color = "black", .facet_filter = {{ .facet_filter }},
+    .max_facets = .max_facets
   )
 
   # Now, prepare matching facets in the predictions
-  ek <- epi_keys_only(plot_data)
+  ek <- epi_keys_only(observed_response)
   predictions <- predictions %>%
     mutate(
       .facets = interaction(!!!rlang::syms(as.list(ek)), sep = " / "),
@@ -215,7 +218,7 @@ autoplot.epi_workflow <- function(
 #' @export
 #' @rdname autoplot-epipred
 autoplot.canned_epipred <- function(
-    object, plot_data = NULL, ...,
+    object, observed_response = NULL, ...,
     .color_by = c("all_keys", "geo_value", "other_keys", ".response", "all", "none"),
     .facet_by = c(".response", "other_keys", "all_keys", "geo_value", "all", "none"),
     .base_color = "dodgerblue4",
@@ -230,9 +233,10 @@ autoplot.canned_epipred <- function(
   predictions <- object$predictions %>%
     rename(time_value = target_date)
 
-  autoplot(ewf, predictions, plot_data, ...,
+  autoplot(ewf, predictions, observed_response, ...,
     .color_by = .color_by, .facet_by = .facet_by,
-    .base_color = .base_color, .facet_filter = {{ .facet_filter }}
+    .base_color = .base_color, .facet_filter = {{ .facet_filter }},
+    .max_facets = .max_facets
   )
 }
 

README.Rmd

@@ -124,7 +124,7 @@ library(ggplot2)
 ```
 </details>
 
-To demonstrate using [`{epipredict}`](https://cmu-delphi.github.io/epipredict/) for forecasting, say we want to
+To demonstrate using [`{epipredict}`](https://cmu-delphi.github.io/epipredict/) for forecasting, suppose we want to
 predict COVID-19 deaths per 100k people for each of a subset of states
 
 ```{r subset_geos}
@@ -226,8 +226,8 @@ cases_deaths <-
 ```
 </details>
 
-After downloading and cleaning the cases and deaths data, we can plot
-a subset of the states, marking the desired forecast date:
+After downloading and cleaning deaths per capita, as well as cases per 100k people, we can plot
+a subset of the states, marking the desired forecast date with a vertical line:
 
 <details>
 <summary> Plot </summary>
@@ -287,16 +287,16 @@ four_week_ahead <- arx_forecaster(
 four_week_ahead
 ```
 
-In our model setup, we are defining as predictors case rate lagged 0-3
-days, one week, and two weeks, and death rate lagged 0-2 weeks.
+In our model setup, we are using as predictors the case rate lagged 0-3
+days, one week, and two weeks, and the death rate lagged 0-2 weeks.
 The result `four_week_ahead` is both a fitted model object which could be used
 any time in the future to create different forecasts, and a set of predicted
 values (and prediction intervals) for each location 28 days after the forecast
 date.
 
 Plotting the prediction intervals on the true values for our location subset[^2]:
 
-[^2]: Alternatively, you could call `autoplot(four_week_ahead, plot_data =
+[^2]: Alternatively, you could call `autoplot(four_week_ahead, observed_response =
     cases_deaths)` to get the full collection of forecasts. This is too busy for
     the space we have for plotting here.
 
@@ -310,14 +310,14 @@ restricted_predictions <-
   mutate(.response_name = "death_rate")
 forecast_plot <-
   four_week_ahead |>
-  autoplot(plot_data = cases_deaths) +
+  autoplot(observed_response = cases_deaths) +
   geom_vline(aes(xintercept = forecast_date)) +
   geom_text(
     data = forecast_date_label %>% filter(.response_name == "death_rate"),
     aes(x = dates, label = "forecast\ndate", y = heights),
     size = 3, hjust = "right"
   ) +
-  scale_x_date(date_breaks = "3 months", date_labels = "%Y %b") +
+  scale_x_date(date_breaks = "3 months", date_labels = "%y %b") +
   theme(axis.text.x = element_text(angle = 90, hjust = 1))
 ```
 </details>

README.md

@@ -62,8 +62,8 @@ library(ggplot2)
 
 To demonstrate using
 [`{epipredict}`](https://cmu-delphi.github.io/epipredict/) for
-forecasting, say we want to predict COVID-19 deaths per 100k people for
-each of a subset of states
+forecasting, suppose we want to predict COVID-19 deaths per 100k people
+for each of a subset of states
 
 ``` r
 used_locations <- c("ca", "ma", "ny", "tx")
@@ -173,8 +173,9 @@ cases_deaths <-
 
 </details>
 
-After downloading and cleaning the cases and deaths data, we can plot a
-subset of the states, marking the desired forecast date:
+After downloading and cleaning deaths per capita, as well as cases per
+100k people, we can plot a subset of the states, marking the desired
+forecast date with a vertical line:
 
 <details>
 <summary>
@@ -234,7 +235,7 @@ four_week_ahead <- arx_forecaster(
 four_week_ahead
 #> ══ A basic forecaster of type ARX Forecaster ════════════════════════════════
 #> 
-#> This forecaster was fit on 2025-03-03 14:43:07.
+#> This forecaster was fit on 2025-04-09 17:23:00.
 #> 
 #> Training data was an <epi_df> with:
 #> • Geography: state,
@@ -251,8 +252,8 @@ four_week_ahead
 #> 
 ```
 
-In our model setup, we are defining as predictors case rate lagged 0-3
-days, one week, and two weeks, and death rate lagged 0-2 weeks. The
+In our model setup, we are using as predictors the case rate lagged 0-3
+days, one week, and two weeks, and the death rate lagged 0-2 weeks. The
 result `four_week_ahead` is both a fitted model object which could be
 used any time in the future to create different forecasts, and a set of
 predicted values (and prediction intervals) for each location 28 days
@@ -274,14 +275,14 @@ restricted_predictions <-
   mutate(.response_name = "death_rate")
 forecast_plot <-
   four_week_ahead |>
-  autoplot(plot_data = cases_deaths) +
+  autoplot(observed_response = cases_deaths) +
   geom_vline(aes(xintercept = forecast_date)) +
   geom_text(
     data = forecast_date_label %>% filter(.response_name == "death_rate"),
     aes(x = dates, label = "forecast\ndate", y = heights),
     size = 3, hjust = "right"
   ) +
-  scale_x_date(date_breaks = "3 months", date_labels = "%Y %b") +
+  scale_x_date(date_breaks = "3 months", date_labels = "%y %b") +
   theme(axis.text.x = element_text(angle = 90, hjust = 1))
 ```
 
@@ -337,9 +338,9 @@ email, or the InsightNet Slack.
     ago.
 
 [^2]: Alternatively, you could call
-    `autoplot(four_week_ahead, plot_data = cases_deaths)` to get the
-    full collection of forecasts. This is too busy for the space we have
-    for plotting here.
+    `autoplot(four_week_ahead, observed_response = cases_deaths)` to get
+    the full collection of forecasts. This is too busy for the space we
+    have for plotting here.
 
 [^3]: Note that these are not the same quantiles that we fit when
     creating `four_week_ahead`. They are extrapolated from those

man/autoplot-epipred.Rd

@@ -337,7 +337,8 @@ p1 <- # first plotting the forecasts as bands, lines and points
   geom_vline(
     data = percent_cli_data |> filter(geo_value == geo_choose) |> select(-version_faithful),
     aes(color = factor(version), xintercept = version),
-    lty = 2) +
+    lty = 2
+  ) +
   # the underlying data
   geom_line(
     data = plotting_data |> filter(geo_value == geo_choose),