Working on optimization

README.Rmd

@@ -434,7 +434,7 @@ sudden there were 72 instead of 70 artworks. It seems like these two
 artworks appear on October 21, 2022.
 
 ```{r}
-dat0 <- read.table("../data/rawdata_logfiles.csv", sep = ";", header = TRUE)
+dat0 <- read.table("data/rawdata_logfiles.csv", sep = ";", header = TRUE)
 dat0$date <- as.POSIXct(dat0$date)
 dat0$glossar <- ifelse(dat0$artwork == "glossar", 1, 0)
 
@@ -448,7 +448,7 @@ summary(dat[dat$artwork %in% c("504", "505"), ])
 The artworks seem to be have updated in general after October 21, 2022.
 
 ```{r}
-art_before_oct2022 <- sort(unique(dat[dat$date >= "2022-10-21", "artwork"]))
+art_after_oct2022 <- sort(unique(dat[dat$date >= "2022-10-21", "artwork"]))
 art_before_oct2022 <- sort(unique(dat[dat$date <= "2022-10-21", "artwork"]))
 # Removed artworks
 art_before_oct2022[!art_before_oct2022 %in% art_after_oct2022]

code/02_preprocessing.R

@@ -13,33 +13,38 @@ dat0$glossar <- ifelse(dat0$artwork == "glossar", 1, 0)
 dat <- subset(dat0, !(dat0$event %in% c("Start Application",
                                         "Show Application")))
 
-save(dat, file = "tmp/dat.RData")
+#save(dat, file = "tmp/dat.RData")
 
 # Add trace variable #####################################################
 cat("########## Adding trace variable... ##########", "\n")
 dat1 <- add_trace(dat)
 
-save(dat1, file = "tmp/dat1.RData")
+#save(dat1, file = "tmp/dat1.RData")
 
 # Close events
-cat("########## Closing events...")
+cat("########## Closing events... ##########", "\n")
-dat2 <- rbind(close_events(dat1, "move"),
+c1 <- close_events(dat1, "move")
-              close_events(dat1, "flipCard"),
+cat("## --> move events closed.", "\n")
-              close_events(dat1, "openTopic"),
+c2 <- close_events(dat1, "flipCard")
-              close_events(dat1, "openPopup"))
+cat("## --> flipCard events closed.", "\n")
+c3 <- close_events(dat1, "openTopic")
+cat("## --> openTopic events closed.", "\n")
+c4 <- close_events(dat1, "openPopup")
+cat("## --> openPopup events closed.", "\n")
+dat2 <- rbind(c1, c2, c3, c4)
 
 dat2 <- dat2[order(dat2$date.start, dat2$fileId.start), ]
 
 # Remove durations when event spans more than one log file, since they are
 # not interpretable
-dat2[which(dat2$fileId.start != dat2$fileId.stop), "duration"] <- NA
+#dat2[which(dat2$fileId.start != dat2$fileId.stop), "duration"] <- NA
 
 # Remove all events that do not have a `date.start`
 dat2 <- dat2[!is.na(dat2$date.start), ]
 rownames(dat2) <- NULL
 # TODO: Throw warning about this
 
-save(dat2, file = "tmp/dat2.RData")
+#save(dat2, file = "tmp/dat2.RData")
 
 # Add case variable ######################################################
 cat("########## Adding case and eventId variables... ##########", "\n")
@@ -56,13 +61,13 @@ dat3 <- dat3[, c("fileId.start", "fileId.stop", "eventId", "case",
                  "scaleSize", "rotation.start", "rotation.stop",
                  "rotationDegree")]
 
-save(dat3, file = "tmp/dat3.RData")
+#save(dat3, file = "tmp/dat3.RData")
 
 # Add trace for move events ##############################################
 cat("########## Adding trace variable for move events... ##########", "\n")
 dat4 <- add_trace_moves(dat3)
 
-save(dat4, file = "tmp/dat4.RData")
+#save(dat4, file = "tmp/dat4.RData")
 
 # Add topics: file names and topics ######################################
 cat("########## Adding information about topics... ##########", "\n")
@@ -72,7 +77,7 @@ topics <- extract_topics(artworks, pattern = paste0(artworks, ".xml"),
 
 dat5 <- add_topic(dat4, topics = topics)
 
-save(dat5, file = "tmp/dat5.RData")
+#save(dat5, file = "tmp/dat5.RData")
 
 # TODO: Replace artwork with informative strings
 

code/functions.R

@@ -147,6 +147,7 @@ close_events <- function(data, event = c("move", "flipCard", "openTopic", "openP
     stop("`event` must be one of 'move', 'flipCard', 'openTopic',
          'openPopup'.")
   }
+# TODO: `fileId` should now maybe go back into `idvar`
 
   subdata <- subset(data, data$event %in% actions)
   #subdata <- subdata[order(subdata$artwork, subdata$popup, subdata$date), ]
@@ -241,9 +242,9 @@ add_case <- function(data, cutoff = 20) {
 # TODO: What is the best choice for the cutoff here?
 
   data$timediff <- as.numeric(diff(c(data$date.start[1], data$date.start)))
-
   data$case <- NA
   j <- 1
+  pb <- txtProgressBar(min = 0, max = nrow(data), style = 3)
 
   for (i in seq_len(nrow(data))) {
     if (data$timediff[i] <= cutoff) {
@@ -252,6 +253,7 @@ add_case <- function(data, cutoff = 20) {
       j <- j + 1
       data$case[i] <- j
     }
+    setTxtProgressBar(pb, i)
   }
   data$timediff <- NULL
   data