feat: add vignette snapshots

Author: dshemetov

tests/testthat/_snaps/vignette-snapshot/advanced.html

@@ -0,0 +1,18 @@
+# Vignettes that use epi_archives or epi_dfs.
+vignettes <- paste0(here::here("vignettes/"), c(
+  "advanced.Rmd",
+  "aggregation.Rmd",
+  "archive.Rmd",
+  "epiprocess.Rmd",
+  "slide.Rmd"
+))
+for (input_file in vignettes) {
+  test_that(paste0("snapshot vignette ", basename(input_file)), {
+    # skip("Skipping snapshot tests by default, as they are slow.")
+    output_file <- sub("\\.Rmd$", ".html", input_file)
+    withr::with_file(output_file, {
+      devtools::build_rmd(input_file)
+      expect_snapshot_file(output_file)
+    })
+  })
+}

tests/testthat/_snaps/vignette-snapshot/advanced.new.html

@@ -9,14 +9,13 @@ vignette: >
 
 In this vignette, we discuss how to use the sliding functionality in the
 `epiprocess` package with less common grouping schemes or with computations that
-have advanced output structures.
-The output of a slide computation should either be an atomic value/vector, or a
-data frame. This data frame can have multiple columns, multiple rows, or both.
+have advanced output structures. The output of a slide computation should either
+be an atomic value/vector, or a data frame. This data frame can have multiple
+columns, multiple rows, or both.
 
 During basic usage (e.g., when all optional arguments are set to their defaults):
 
 * `epi_slide(edf, <computation>, .....)`:
-
   * keeps **all** columns of `edf`, adds computed column(s)
   * outputs **one row per row in `edf`** (recycling outputs from
     computations appropriately if there are multiple time series bundled
@@ -26,9 +25,7 @@ During basic usage (e.g., when all optional arguments are set to their defaults)
     `dplyr::arrange(time_value, .by_group = TRUE)`**
   * outputs an **`epi_df`** if the required columns are present, otherwise a
     tibble
-
 * `epix_slide(ea, <computation>, .....)`:
-
   * keeps **grouping and `time_value`** columns of `ea`, adds computed
     column(s)
   * outputs **any number of rows** (computations are allowed to output any
@@ -40,6 +37,7 @@ During basic usage (e.g., when all optional arguments are set to their defaults)
   * outputs a **tibble**
 
 These differences in basic behavior make some common slide operations require less boilerplate:
+
 * predictors and targets calculated with `epi_slide` are automatically lined up
   with each other and with the signals from which they were calculated; and
 * computations for an `epix_slide` can output data frames with any number of
@@ -84,13 +82,14 @@ simple synthetic example.
 ```{r message = FALSE}
 library(epiprocess)
 library(dplyr)
+set.seed(123)
 
 edf <- tibble(
   geo_value = rep(c("ca", "fl", "pa"), each = 3),
   time_value = rep(seq(as.Date("2020-06-01"), as.Date("2020-06-03"), by = "day"), length.out = length(geo_value)),
   x = seq_along(geo_value) + 0.01 * rnorm(length(geo_value)),
 ) %>%
-  as_epi_df()
+  as_epi_df(as_of = as.Date("2024-03-20"))
 
 # 2-day trailing average, per geo value
 edf %>%
@@ -338,7 +337,7 @@ library(data.table)
 library(ggplot2)
 theme_set(theme_bw())
 
-x <- archive_cases_dv_subset_2$DT %>%
+x <- archive_cases_dv_subset$DT %>%
   filter(geo_value %in% c("ca", "fl")) %>%
   as_epi_archive(compactify = FALSE)
 ```
@@ -525,10 +524,7 @@ separate ARX model on each state. As in the archive vignette, we can see a
 difference between version-aware (right column) and -unaware (left column)
 forecasting, as well.
 
-
 ## Attribution
 The `case_rate_7d_av` data used in this document is a modified part of the [COVID-19 Data Repository by the Center for Systems Science and Engineering (CSSE) at Johns Hopkins University](https://github.com/CSSEGISandData/COVID-19) as [republished in the COVIDcast Epidata API](https://cmu-delphi.github.io/delphi-epidata/api/covidcast-signals/jhu-csse.html). This data set is licensed under the terms of the [Creative Commons Attribution 4.0 International license](https://creativecommons.org/licenses/by/4.0/) by the Johns Hopkins University on behalf of its Center for Systems Science in Engineering. Copyright Johns Hopkins University 2020.
 
 The `percent_cli` data is a modified part of the [COVIDcast Epidata API Doctor Visits data](https://cmu-delphi.github.io/delphi-epidata/api/covidcast-signals/doctor-visits.html). This dataset is licensed under the terms of the [Creative Commons Attribution 4.0 International license](https://creativecommons.org/licenses/by/4.0/). Copyright Delphi Research Group at Carnegie Mellon University 2020.
-
-

tests/testthat/_snaps/vignette-snapshot/aggregation.html

@@ -34,7 +34,7 @@ x <- pub_covidcast(
 ) %>%
   select(geo_value, time_value, cases = value) %>%
   full_join(y, by = "geo_value") %>%
-  as_epi_df()
+  as_epi_df(as_of = as.Date("2024-03-20"))
 ```
 
 The data contains 16,212 rows and 5 columns.
@@ -192,7 +192,7 @@ running `epi_slide()` on the zero-filled data brings these trailing averages
 
 ```{r}
 xt %>%
-  as_epi_df() %>%
+  as_epi_df(as_of = as.Date("2024-03-20")) %>%
   group_by(geo_value) %>%
   epi_slide(cases_7dav = mean(cases), before = 6) %>%
   ungroup() %>%
@@ -203,7 +203,7 @@ xt %>%
   print(n = 7)
 
 xt_filled %>%
-  as_epi_df() %>%
+  as_epi_df(as_of = as.Date("2024-03-20")) %>%
   group_by(geo_value) %>%
   epi_slide(cases_7dav = mean(cases), before = 6) %>%
   ungroup() %>%