doc: update vignettes

Author: dshemetov

.Rbuildignore

@@ -16,3 +16,5 @@
 ^.lintr$
 ^DEVELOPMENT.md$
 man-roxygen
+^.venv$
+^sandbox.R$

.gitignore

@@ -13,3 +13,4 @@ docs
 renv/
 renv.lock
 .Rprofile
+sandbox.R

DESCRIPTION

@@ -50,7 +50,8 @@ Imports:
     tidyselect (>= 1.2.0),
     tsibble,
     utils,
-    vctrs
+    vctrs,
+    waldo
 Suggests:
     covidcast,
     devtools,

R/methods-epi_archive.R

@@ -731,7 +731,7 @@ epix_detailed_restricted_mutate <- function(.data, ...) {
 #' library(dplyr)
 #'
 #' # Reference time points for which we want to compute slide values:
-#' versions <- seq(as.Date("2020-06-01"),
+#' versions <- seq(as.Date("2020-06-02"),
 #'   as.Date("2020-06-15"),
 #'   by = "1 day"
 #' )
@@ -780,7 +780,7 @@ epix_detailed_restricted_mutate <- function(.data, ...) {
 #'     .versions = versions
 #'   ) %>%
 #'   ungroup() %>%
-#'   arrange(geo_value, time_value)
+#'   arrange(geo_value, version)
 #'
 #' # --- Advanced: ---
 #'

R/methods-epi_df.R

@@ -427,6 +427,10 @@ arrange_col_canonical.epi_df <- function(x, ...) {
   x %>% dplyr::relocate(dplyr::all_of(cols), .before = 1)
 }
 
+#' Group an `epi_df` object by default keys
+#' @param x an `epi_df`
+#' @param exclude character vector of column names to exclude from grouping
+#' @return a grouped `epi_df`
 #' @export
 group_epi_df <- function(x, exclude = character()) {
   cols <- key_colnames(x, exclude = exclude)
@@ -440,7 +444,7 @@ group_epi_df <- function(x, exclude = character()) {
 #' the resulting `epi_df` will have `geo_value` set to `"total"`.
 #'
 #' @param .x an `epi_df`
-#' @param value_col character vector of the columns to aggregate
+#' @param sum_cols character vector of the columns to aggregate
 #' @param group_cols character vector of column names to group by. "time_value" is
 #' included by default.
 #' @return an `epi_df` object

R/outliers.R

@@ -161,8 +161,7 @@ detect_outlr <- function(x = seq_along(y), y,
 #'   group_by(geo_value) %>%
 #'   mutate(outlier_info = detect_outlr_rm(
 #'     x = time_value, y = cases
-#'   )) %>%
-#'   unnest(outlier_info)
+#'   ))
 detect_outlr_rm <- function(x = seq_along(y), y, n = 21,
                             log_transform = FALSE,
                             detect_negatives = FALSE,
@@ -189,7 +188,7 @@ detect_outlr_rm <- function(x = seq_along(y), y, n = 21,
 
   # Calculate lower and upper thresholds and replacement value
   z <- z %>%
-    epi_slide(fitted = median(y), .window_size = n, .align = "center") %>%
+    epi_slide(fitted = median(y, na.rm = TRUE), .window_size = n, .align = "center") %>%
     dplyr::mutate(resid = y - fitted) %>%
     roll_iqr(
       n = n,
@@ -256,9 +255,8 @@ detect_outlr_rm <- function(x = seq_along(y), y, n = 21,
 #'   group_by(geo_value) %>%
 #'   mutate(outlier_info = detect_outlr_stl(
 #'     x = time_value, y = cases,
-#'     seasonal_period = 7
-#'   )) %>% # weekly seasonality for daily data
-#'   unnest(outlier_info)
+#'     seasonal_period = 7 # weekly seasonality for daily data
+#'   ))
 detect_outlr_stl <- function(x = seq_along(y), y,
                              n_trend = 21,
                              n_seasonal = 21,
@@ -359,7 +357,7 @@ roll_iqr <- function(z, n, detection_multiplier, min_radius,
 
   z %>%
     epi_slide(
-      roll_iqr = stats::IQR(resid),
+      roll_iqr = stats::IQR(resid, na.rm = TRUE),
       .window_size = n, .align = "center"
     ) %>%
     dplyr::mutate(

R/revision_analysis.R

@@ -81,8 +81,8 @@ revision_summary <- function(epi_arch,
                              should_compactify = TRUE) {
   arg <- names(eval_select(rlang::expr(c(...)), allow_rename = FALSE, data = epi_arch$DT))
   if (length(arg) == 0) {
-    first_non_key <- !(names(epi_arch$DT) %in% c(key_colnames(epi_arch), "version"))
-    arg <- names(epi_arch$DT)[first_non_key][1]
+    # Choose the first column that's not a key or version
+    arg <- setdiff(names(epi_arch$DT), c(key_colnames(epi_arch), "version"))[[1]]
   } else if (length(arg) > 1) {
     cli_abort("Not currently implementing more than one column at a time. Run each separately")
   }
@@ -99,11 +99,9 @@ revision_summary <- function(epi_arch,
   #
   # revision_tibble
   keys <- key_colnames(epi_arch)
-  names(epi_arch$DT)
 
-  revision_behavior <-
-    epi_arch$DT %>%
-    select(c(geo_value, time_value, all_of(keys), version, !!arg))
+  revision_behavior <- epi_arch$DT %>%
+    select(all_of(unique(c("geo_value", "time_value", keys, "version", arg))))
   if (!is.null(min_waiting_period)) {
     revision_behavior <- revision_behavior %>%
       filter(abs(time_value - as.Date(epi_arch$versions_end)) >= min_waiting_period)

R/slide.R

@@ -728,7 +728,7 @@ epi_slide_opt <- function(
   # positions of user-provided `col_names` into string column names. We avoid
   # using `names(pos)` directly for robustness and in case we later want to
   # allow users to rename fields via tidyselection.
-  if (class(quo_get_expr(enquo(.col_names))) == "character") {
+  if (inherits(quo_get_expr(enquo(.col_names)), "character")) {
     pos <- eval_select(dplyr::all_of(.col_names), data = .x, allow_rename = FALSE)
   } else {
     pos <- eval_select(enquo(.col_names), data = .x, allow_rename = FALSE)

_pkgdown.yml

@@ -48,7 +48,6 @@ articles:
       - aggregation
       - outliers
       - archive
-      - advanced
       - compactify
 
 repo:

man-roxygen/basic-slide-details.R

@@ -9,7 +9,7 @@
 #'   boundary of the dataset) and will attempt to perform the computation
 #'   anyway. The issue of what to do with partial computations (those run on
 #'   incomplete windows) is therefore left up to the user, either through the
-#'   specified function or formula `f`, or through post-processing.
+#'   specified function or formula, or through post-processing.
 #'
 #'   Let's look at some window examples, assuming that the reference time value
 #'   is "tv". With .align = "right" and .window_size = 3, the window will be: