Dev

R/simaerep.R

@@ -448,15 +448,20 @@ eval_sites <- function(df_sim_sites,
 #'@keywords internal
 p_adjust <- function(df, col, suffix, method = "BH") {
 
+  col_adj <- paste0(col, "_adj")
+  col_suffix <- paste0(col, suffix)
+
   if (is.na(method) || is.null(method) || method %in% c("None", "none")) {
-    return(df)
+    df_out <- df %>%
+      mutate(
+        !! as.name(col_suffix) := 1 - .data[[col]]
+      )
+
+    return(df_out)
   }
 
   if (inherits(df, "data.frame")) {
 
-  col_adj <- paste0(col, "_adj")
-  col_suffix <- paste0(col, suffix)
-
   df_out <- df %>%
     mutate(
       !! as.name(col_adj) := p.adjust(.data[[col]], method = method),

R/simaerep_plot.R

@@ -539,14 +539,26 @@ plot_study <- function(df_visit,
 
   # ordered sites -------------------------------------------------------------
 
+  if (any(stringr::str_detect(colnames(df_eval), "_adj$"))) {
+    thresh <- 0.5
+    breaks <- c(0, 0.5, 0.75, 0.95, ifelse(max(df_eval[[prob_col]], na.rm = TRUE) > 0.95,
+                                           max(df_eval[[prob_col]], na.rm = TRUE) + 0.1,
+                                           NA))
+    } else {
+    thresh <- 0.9
+    breaks <- c(0, 0.9, 0.95, 0.99, ifelse(max(df_eval[[prob_col]], na.rm = TRUE) > 0.95,
+                                           max(df_eval[[prob_col]], na.rm = TRUE) + 0.1,
+                                           NA))
+    }
+
   n_site_ur_gr_0p5 <- df_eval %>%
-    filter(.data[[prob_col]] > 0.5) %>%
+    filter(.data[[prob_col]] > thresh) %>%
     nrow()
 
   if (n_site_ur_gr_0p5 > 0) {
     sites_ordered <- df_eval %>%
       arrange(.data$study_id, desc(.data[[prob_col]]), .data[[col_mean_site]]) %>%
-      filter(.data[[prob_col]] > 0.5) %>%
+      filter(.data[[prob_col]] > thresh) %>%
       head(n_sites) %>%
       .$site_number
   } else {
@@ -604,10 +616,7 @@ plot_study <- function(df_visit,
   # define score cut-offs + labels----------------------------------------------
 
   palette <- RColorBrewer::brewer.pal(9, "Blues")[c(3, 5, 7, 9)]
-  breaks <- c(0, 0.5, 0.75, 0.95, ifelse(max(df_eval[[prob_col]], na.rm = TRUE) > 0.95,
-                                    max(df_eval[[prob_col]], na.rm = TRUE) + 0.1,
-                                    NA)
-              )
+
 
   breaks <- breaks[! is.na(breaks)]
 

README.Rmd

@@ -71,6 +71,9 @@ Download as pdf in the [release section](https://github.com/openpharma/simaerep/
 
 ## Application
 
+Recommended Threshold: `aerep$dfeval$prob_low_prob_ur: 0.95`
+
+
 ```{r fig.width=10}
 
 suppressPackageStartupMessages(library(simaerep))
@@ -96,9 +99,57 @@ df_visit %>%
 
 aerep <- simaerep(df_visit)
 
-plot(aerep, study = "A") 
+plot(aerep, study = "A")
 ```
 
 *Left panel shows mean AE reporting per site (lightblue and darkblue lines) against mean AE reporting of the entire study (golden line). Single sites are plotted in descending order by AE under-reporting probability on the right panel in which grey lines denote cumulative AE count of single patients. Grey dots in the left panel plot indicate sites that were picked for single plotting. AE under-reporting probability of dark blue lines crossed threshold of 95%. Numbers in the upper left corner indicate the ratio of patients that have been used for the analysis against the total number of patients. Patients that have not been on the study long enough to reach the evaluation point (visit_med75, see introduction) will be ignored.*
 
+## Optimized Statistical Performance
+
+Following the recommendation of our latest [performance benchmark](https://openpharma.github.io/simaerep/articles/performance.html) statistical performance can be increased by using the [inframe](https://openpharma.github.io/simaerep/articles/inframe.html) algorithm without multiplicity correction.
+
+**Note that the plot is more noisy because no patients are excluded and only a few patients contribute to the event count at higher visits**
+
+Recommended Threshold: `aerep$dfeval$prob_low_prob_ur: 0.99`
+
+```{r}
+aerep <- simaerep(
+  df_visit,
+  inframe = TRUE,
+  visit_med75 = FALSE,
+  mult_corr = FALSE
+)
+
+plot(aerep, study = "A")
+```
+
+## In Database Calculation
+
+The [inframe](https://openpharma.github.io/simaerep/articles/inframe.html) algorithm uses only `dbplyr` compatible table operations and can be executed within a database backend as we demonstrate here using `duckdb`.
+
+However, we need to provide a in database table that has as many rows as the desired replications in our simulation, instead of providing an integer for the `r` parameter.
+
+```{r}
+con <- DBI::dbConnect(duckdb::duckdb(), dbdir = ":memory:")
+df_r <- tibble(rep = seq(1, 1000))
+
+dplyr::copy_to(con, df_visit, "visit")
+dplyr::copy_to(con, df_r, "r")
+
+tbl_visit <- tbl(con, "visit")
+tbl_r <- tbl(con, "r")
+
+
+aerep <- simaerep(
+  tbl_visit,
+  r = tbl_r,
+  inframe = TRUE,
+  visit_med75 = FALSE,
+  mult_corr = FALSE
+)
+
+plot(aerep, df_visit = tbl_visit)
+
+DBI::dbDisconnect(con)
+```
 

README.md

@@ -87,6 +87,8 @@ using
 
 ## Application
 
+Recommended Threshold: `aerep$dfeval$prob_low_prob_ur: 0.95`
+
 ``` r
 
 suppressPackageStartupMessages(library(simaerep))
@@ -143,7 +145,7 @@ df_visit %>%
 
 aerep <- simaerep(df_visit)
 
-plot(aerep, study = "A") 
+plot(aerep, study = "A")
 ```
 
 <img src="man/figures/README-unnamed-chunk-2-1.png" width="100%" />
@@ -159,3 +161,71 @@ upper left corner indicate the ratio of patients that have been used for
 the analysis against the total number of patients. Patients that have
 not been on the study long enough to reach the evaluation point
 (visit_med75, see introduction) will be ignored.*
+
+## Optimized Statistical Performance
+
+Following the recommendation of our latest [performance
+benchmark](https://openpharma.github.io/simaerep/articles/performance.html)
+statistical performance can be increased by using the
+[inframe](https://openpharma.github.io/simaerep/articles/inframe.html)
+algorithm without multiplicity correction.
+
+**Note that the plot is more noisy because no patients are excluded and
+only a few patients contribute to the event count at higher visits**
+
+Recommended Threshold: `aerep$dfeval$prob_low_prob_ur: 0.99`
+
+``` r
+aerep <- simaerep(
+  df_visit,
+  inframe = TRUE,
+  visit_med75 = FALSE,
+  mult_corr = FALSE
+)
+
+plot(aerep, study = "A")
+```
+
+<img src="man/figures/README-unnamed-chunk-3-1.png" width="100%" />
+
+## In Database Calculation
+
+The
+[inframe](https://openpharma.github.io/simaerep/articles/inframe.html)
+algorithm uses only `dbplyr` compatible table operations and can be
+executed within a database backend as we demonstrate here using
+`duckdb`.
+
+However, we need to provide a in database table that has as many rows as
+the desired replications in our simulation, instead of providing an
+integer for the `r` parameter.
+
+``` r
+con <- DBI::dbConnect(duckdb::duckdb(), dbdir = ":memory:")
+df_r <- tibble(rep = seq(1, 1000))
+
+dplyr::copy_to(con, df_visit, "visit")
+dplyr::copy_to(con, df_r, "r")
+
+tbl_visit <- tbl(con, "visit")
+tbl_r <- tbl(con, "r")
+
+
+aerep <- simaerep(
+  tbl_visit,
+  r = tbl_r,
+  inframe = TRUE,
+  visit_med75 = FALSE,
+  mult_corr = FALSE
+)
+
+plot(aerep, df_visit = tbl_visit)
+#> study = NULL, defaulting to study:A
+```
+
+<img src="man/figures/README-unnamed-chunk-4-1.png" width="100%" />
+
+``` r
+
+DBI::dbDisconnect(con)
+```

cran-comments.md

@@ -1,12 +1,12 @@
-# submission simaerep v0.5.0
+# submission simaerep v0.6.0
 
 ## Test Results
 
 No notes, warnings or errors
 
 ## Test Environments
 
-- Rhub, debian
+- Rhub - linux, macos, macos-arm64, windows devel
 - GitHub CI/CD macOS-latest