From fd4e47bd29debea44a5ad0c3fbbab2896b96df0c Mon Sep 17 00:00:00 2001
From: Aurelien Dugourd <dugourd@ebi.ac.uk>
Date: Tue, 3 Dec 2024 11:18:14 +0000
Subject: [PATCH] some test from scratch and visual enhancement to existing

---
 .gitignore                        |  7 +++
 F1000_Slack.Rproj                 | 17 +++++++
 aurelien_sciwheel_stats_process.R | 74 +++++++++++++++++++++++++++++++
 sciwheel_stats_process.R          |  6 +--
 4 files changed, 101 insertions(+), 3 deletions(-)
 create mode 100644 F1000_Slack.Rproj
 create mode 100644 aurelien_sciwheel_stats_process.R

diff --git a/.gitignore b/.gitignore
index 53d0fcb..a0195ff 100644
--- a/.gitignore
+++ b/.gitignore
@@ -19,6 +19,7 @@
 
 # RStudio files
 .Rproj.user/
+.Rbuildignore
 
 # produced vignettes
 vignettes/*.html
@@ -38,3 +39,9 @@ vignettes/*.pdf
 # R Environment Variables
 .Renviron
 *.rdata
+
+# data object
+Sciwheel_update/**
+
+# plots
+overviewplots**
diff --git a/F1000_Slack.Rproj b/F1000_Slack.Rproj
new file mode 100644
index 0000000..21a4da0
--- /dev/null
+++ b/F1000_Slack.Rproj
@@ -0,0 +1,17 @@
+Version: 1.0
+
+RestoreWorkspace: Default
+SaveWorkspace: Default
+AlwaysSaveHistory: Default
+
+EnableCodeIndexing: Yes
+UseSpacesForTab: Yes
+NumSpacesForTab: 2
+Encoding: UTF-8
+
+RnwWeave: Sweave
+LaTeX: pdfLaTeX
+
+BuildType: Package
+PackageUseDevtools: Yes
+PackageInstallArgs: --no-multiarch --with-keep.source
diff --git a/aurelien_sciwheel_stats_process.R b/aurelien_sciwheel_stats_process.R
new file mode 100644
index 0000000..a2c923c
--- /dev/null
+++ b/aurelien_sciwheel_stats_process.R
@@ -0,0 +1,74 @@
+nested_list <- readRDS("data_dump.rds")
+
+# Extracting "f1000AddedBy" field
+extract_added_by <- function(list_item) {
+  sapply(list_item$content, function(content_item) content_item$f1000AddedBy)
+}
+
+# Flatten the list of "f1000AddedBy" values
+added_by_list <- unlist(lapply(nested_list, extract_added_by), use.names = FALSE)
+
+# Count occurrences of each lab member
+library(dplyr)
+added_by_count <- as.data.frame(table(added_by_list)) %>%
+  rename(Member = added_by_list, Count = Freq)
+
+# Plot the data
+library(ggplot2)
+ggplot(added_by_count, aes(x = Member, y = Count)) +
+  geom_bar(stat = "identity") +
+  labs(title = "Number of Papers Added by Each Lab Member",
+       x = "Lab Member",
+       y = "Number of Papers") +
+  theme_minimal()
+
+
+# Sort the data in descending order of counts
+added_by_count <- added_by_count %>%
+  arrange(desc(Count))
+
+# Plot the data with horizontal bars and sorted order
+ggplot(added_by_count, aes(x = reorder(Member, Count), y = Count)) +
+  geom_bar(stat = "identity") +
+  labs(title = "Number of Papers Added by Each Lab Member",
+       x = "Lab Member",
+       y = "Number of Papers") +
+  theme_minimal() +
+  coord_flip()
+
+
+#####
+
+# Extracting "f1000AddedBy" and "f1000AddedDate" fields robustly
+extract_added_info <- function(list_item) {
+  if (is.null(list_item$content)) return(data.frame(f1000AddedBy = NA, f1000AddedDate = NA))
+  do.call(rbind, lapply(list_item$content, function(content_item) {
+    data.frame(
+      f1000AddedBy = ifelse(is.null(content_item$f1000AddedBy), NA, content_item$f1000AddedBy),
+      f1000AddedDate = ifelse(is.null(content_item$f1000AddedDate), NA, content_item$f1000AddedDate)
+    )
+  }))
+}
+
+# Flatten and transform the nested list into a single data frame
+added_info <- do.call(rbind, lapply(nested_list, extract_added_info))
+
+# Convert dates from millis to R date format and filter by year 2024
+added_info <- added_info %>%
+  mutate(f1000AddedDate = as.POSIXct(as.numeric(f1000AddedDate) / 1000, origin = "1970-01-01"),
+         Year = format(f1000AddedDate, "%Y")) %>%
+  filter(Year == "2024" & !is.na(f1000AddedBy))
+
+# Count occurrences of each lab member
+added_by_count <- added_info %>%
+  count(f1000AddedBy, name = "Count") %>%
+  rename(Member = f1000AddedBy)
+
+# Plot the data with horizontal bars and sorted order
+ggplot(added_by_count, aes(x = reorder(Member, Count), y = Count)) +
+  geom_bar(stat = "identity") +
+  labs(title = "Number of Papers Added by Each Lab Member in 2024",
+       x = "Lab Member",
+       y = "Number of Papers") +
+  theme_minimal() +
+  coord_flip()
diff --git a/sciwheel_stats_process.R b/sciwheel_stats_process.R
index 49765c0..69c44ed 100644
--- a/sciwheel_stats_process.R
+++ b/sciwheel_stats_process.R
@@ -158,7 +158,7 @@ per.user.df= lapply(names(per.user),function(x) (enframe(per.user[[x]]) %>% muta
 
 #simple plot
 ggplot(per.user.df, aes(x= name, y= folder, fill = value))+
-  geom_tile()
+  geom_tile() + theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
 
 
 # Notes  --------------------------------------------------------------------------------
@@ -233,7 +233,7 @@ tags.per.club.df= lapply(names(tags.per.club),function(x) (enframe(tags.per.club
 
 #simple plot
 ggplot(tags.per.club.df, aes(x= name, y= folder, fill = value))+
-  geom_tile()
+  geom_tile() + theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))
 
 #### tags per user
   tags.per.user = 
@@ -259,7 +259,7 @@ tags.per.user.df= lapply(names(tags.per.user),function(x) (do.call(rbind,tags.pe
 
 #simple plot
 ggplot(tags.per.user.df, aes(x= name, y= value, fill = n))+
-  geom_tile()
+  geom_tile() + theme(axis.text.x = element_text(angle = 45, vjust = 1, hjust=1))