#Loading required libraries library(tidyverse) library(Hmisc) library(stringr) library(reshape2) library(dplyr) library(splitstackshape) library(chron) library(ggplot2) library(tidyverse) library(ggthemes) run_results <- read.table("/home/asiimwe/projects/run_env/alpha_star_wasp_comparison/benchmark_results_all_runs.txt", sep="\t", header = FALSE, fill=TRUE, quote = "") run_results <- as.data.frame(run_results[-c(1),]) run_results$V4 <- NULL colnames(run_results) <- c("Sample", "Run", "Thread", "Param","Value") run_results$Param <- as.character(run_results$Param) run_results$Param[run_results$Param == "Elapsed (wall clock) time (h"] <- "Wall_Clock" run_results$Param[run_results$Param == "Maximum resident set size (kbytes)"] <- "Memory" run_results$Value <- as.character(run_results$Value) run_results$Value <- gsub("mm:ss or m:ss): ", "", run_results$Value) run_results$Value <- gsub(" ", "", run_results$Value) run_results$Value unique(run_results$Run) run_results_STAR <- run_results %>% filter(Run == "STAR_Runs_noVars_noWaspOutput")#STAR base run_results_STAR_WASP <- run_results %>% filter(Run == "STAR_Runs") #STAR WASP run_results_WASP <- run_results %>% filter(Run == "STAR_WASP_Runs") #WASP - change naming globally run_results_STAR$Run <- NULL run_results_STAR_WASP$Run <- NULL run_results_WASP$Run <- NULL run_results_STAR <- spread(run_results_STAR, key = Thread, value = Value) colnames(run_results_STAR) <- c("Sample", "Param", "16threads-STAR", "32threads-STAR", "8threads-STAR") run_results_STAR_WASP <- spread(run_results_STAR_WASP, key = Thread, value = Value) colnames(run_results_STAR_WASP) <- c("Sample", "Param", "16threads-STAR_WASP", "32threads-STAR_WASP", "8threads-STAR_WASP") run_results_WASP <- spread(run_results_WASP, key = Thread, value = Value) colnames(run_results_WASP) <- c("Sample", "Param", "16threads-WASP", "32threads-WASP", "8threads-WASP") data <- merge(run_results_STAR, run_results_STAR_WASP, by = c("Sample", "Param")) data <- merge(data, run_results_WASP, by = c("Sample", "Param")) #write.csv(data, file = "/home/asiimwe/projects/run_env/alpha_star_wasp_comparison/benchmark_results_all_runs_pivoted.csv") #Plotting speeds and memory data_melted <- melt(data, id.vars=c("Sample", "Param")) data_melted <- cSplit(data_melted,"variable", "-", direction="wide") colnames(data_melted) <- c("Sample", "Param", "Value", "Threads", "Run") data_melted$Param <- as.character(data_melted$Param) #data_melted_subset_clock <- data_melted_subset %>% filter(Param == "Wall_Clock") #data_melted_subset_clock$Value <- gsub("\\.", ":", data_melted_subset_clock$Value) #data_melted_subset_clock$Value #data_melted_subset_clock$Value <- chron(times. = data_melted_subset_clock$Value)#rejecting format - expecting h:m:s but third aguments is more than 60 for some instances ###########------------------------------------------------------------------------------ # #compressed dataset - removing parameters that are not needed # #parameters to remove: # param_remove <- c( # "Average resident set size (kbytes)", # "Average shared text size (kbytes)", # "Average stack size (kbytes)", # "Average total size (kbytes)", # "Average unshared data size (kbytes)", # "Exit status", # "Major (requiring I/O) page faults", # "Signals delivered", # "Socket messages received", # "Socket messages sent", # "Swaps" # ) # # dataset_subset <- data %>% filter(Param %nin% param_remove) # write.csv(dataset_subset, file = "/home/asiimwe/projects/run_env/alpha_star_wasp_comparison/run_results_subset_pivoted.csv") ###########------------------------------------------------------------------------------ #Benchmark plots: #Elapsed (wall clock) time (h mm:ss or m:ss): Average resident set size (kbytes) final.log.results <- read.table("/home/asiimwe/projects/run_env/alpha_star_wasp_comparison/final_log_results_all_runs.txt", sep="|", header = FALSE, fill=TRUE, quote = "") colnames(final.log.results) <- unlist(final.log.results[c(1),]) final.log.results <- as.data.frame(final.log.results[-c(1),]) final.log.results <- as.data.frame(final.log.results) final.log.results$Sample <- as.character(final.log.results$Sample) final.log.results$Run <- as.character(final.log.results$Run) final.log.results$Thread <- as.character(final.log.results$Thread) final.log.results$Param <- as.character(final.log.results$Param) colnames(final.log.results) final.log.results$Param <- str_trim(final.log.results$Param) final.log.results_filt <- final.log.results %>% filter(Value != "") final.log.results_filt$Run <- as.character(final.log.results_filt$Run) unique(final.log.results_filt$Run) final.log.results_filt$Run[final.log.results_filt$Run == "STAR_Runs_noVars_noWaspOutput"] <- "STAR" final.log.results_filt$Run[final.log.results_filt$Run == "STAR_Runs"] <- "STAR_WASP" final.log.results_filt$Run[final.log.results_filt$Run == "STAR_WASP_Runs"] <- "WASP" colnames(final.log.results_filt)[3] <- "Threads" #Extracting number of reads for now final.log.results_filt_input_reads <- final.log.results_filt %>% filter(Param == "Number of input reads") final.log.results_filt_mapping_speed <- final.log.results_filt %>% filter(Param == "Mapping speed, Million of reads per hour") final.log.results_filt_average_input_read_length <- final.log.results_filt %>% filter(Param == "Average input read length") data_melted_join_pass1 <- inner_join(data_melted, final.log.results_filt_input_reads, by = c("Sample", "Run", "Threads")) data_melted_join_pass2 <- inner_join(data_melted_join_pass1, final.log.results_filt_mapping_speed, by = c("Sample", "Run", "Threads")) data_melted_beta <- inner_join(data_melted_join_pass2, final.log.results_filt_average_input_read_length, by = c("Sample", "Run", "Threads")) colnames(data_melted_beta) colnames(data_melted_beta) <- c( "Sample", "Param", "Value", "Threads", "Run", "Param.y" , "Number_of_input_reads", "Param.x.x", "Mapping_speed_Million_of_reads_per_hour", "Param.y.y", "Average_input_read_length") data_melted_beta$Param.y <- NULL data_melted_beta$Param.x.x <- NULL data_melted_beta$Param.y.y <- NULL nrow(data_melted) nrow(final.log.results_filt_input_reads) nrow(final.log.results_filt_mapping_speed) nrow(final.log.results_filt_average_input_read_length) nrow(data_melted_beta) #write.csv(data_melted_beta, file = "/home/asiimwe/projects/run_env/alpha_star_wasp_comparison/data_melted_beta.csv") data_melted_beta <- read.csv("/home/asiimwe/projects/run_env/alpha_star_wasp_comparison/data_melted_beta.csv") data_melted_beta$X <- NULL colnames(data_melted_beta) #"Sample" "Run" "Threads" "Param" "Value" "N_Reads" "Speed" #Data cleanup str(data_melted_beta) data_melted_beta$Sample <- as.character(data_melted_beta$Sample) data_melted_beta$Param <- as.character(data_melted_beta$Param) data_melted_beta$Value <- gsub("%", "", data_melted_beta$Value) data_melted_beta %>% filter(Param == "Wall_Clock") #We have different formats for the wall clock - 1:11:33 or 59:38.85 get meanings and convert #We need to reorder Threads to appear in the correct order data_melted_beta$Threads <- ordered(data_melted_beta$Threads , levels = c("8threads", "16threads", "32threads")) dat_text <- data.frame( label = c("8 Threads", "16 Threads", "32 Threads"), Threads = c("8threads", "16threads", "32threads") # Run = factor(3,levels = c(8,16,32)) ) data_melted_beta_memory <- data_melted_beta %>% filter(Param == "Memory") data_melted_beta_memory$Value <- as.numeric(data_melted_beta_memory$Value) data_melted_beta_memory <- data_melted_beta_memory[order(data_melted_beta_memory$Param),] range(data_melted_beta_memory$Value) #30649908 38445368 #Memory data_melted_beta_memory %>% filter(Param == "Memory") %>% ggplot(aes(x = Sample, y = Value, group=Run)) + geom_line(aes(color = Run, linetype = Run)) + scale_color_manual(values = c("darkred", "steelblue", "cyan3")) + facet_wrap(~Threads) + labs(y = "Memory", x="") + theme_light() + theme(legend.title = element_blank()) + #Base, bw, excel_new(), few, light, lindraw #theme(strip.background =element_rect(fill="white"))+ # theme(strip.text = element_text(colour = 'black')) + theme(strip.text.x = element_blank()) + theme(axis.text.x = element_text(angle = 90,hjust = 1,vjust = 0.5, size=10))# + #geom_text(x = 3, y = max(data_melted_beta_memory$Value), aes(label = label), data = dat_text, check_overlap = TRUE) #Mapping Speed data_melted_beta_speed <- data_melted_beta[order(data_melted_beta$Sample, data_melted_beta$Mapping_speed_Million_of_reads_per_hour),] data_melted_beta_speed %>% ggplot(aes(x = Sample, y = Mapping_speed_Million_of_reads_per_hour, group=Run)) + geom_line(aes(color = Run, linetype = Run)) + scale_color_manual(values = c("darkred", "steelblue", "cyan3")) + facet_wrap(~Threads) + labs(y = "Mapping_Speed", x="") + theme_light() + theme(legend.title = element_blank()) + #Base, bw, excel_new(), few, light, lindraw #theme(strip.background =element_rect(fill="white"))+ # theme(strip.text = element_text(colour = 'black')) + theme(strip.text.x = element_blank()) + theme(axis.text.x = element_text(angle = 90,hjust = 1,vjust = 0.5, size=10))# + #geom_text(x = 3, y = max(data_melted_beta_memory$Value), aes(label = label), data = dat_text, check_overlap = TRUE) #put threads, order, look into other stats, tables #Latex installations # tinytex::parse_install( # text = "! LaTeX Error: File `ae.sty' not found." # ) # # # tinytex::parse_install( # text = "! LaTeX Error: File `fullpage.sty' not found." # ) # tinytex::parse_install( # text = "! LaTeX Error: File `moresize.sty' not found." # ) # # # tinytex::parse_install( # text = "! LaTeX Error: File `colortbl.sty' not found." # ) # # # tinytex::parse_install( # text = "! LaTeX Error: File `soul.sty' not found." # ) # # tinytex::parse_install( # text = "! LaTeX Error: File `adjustbox.sty' not found." # ) # # # tinytex::parse_install( # text = "! LaTeX Error: File `collectbox.sty' not found." # ) #Preparing data for latex rep data_melted_beta data_melted_beta$Threads <- ordered(data_melted_beta$Threads , levels = c("8threads", "16threads", "32threads")) unique(data_melted_beta$Param) data_melted_beta_subset <- data_melted_beta %>% filter(Param == "Wall_Clock" | Param == "Memory") data_melted_beta_subset$Average_input_read_length <- NULL #data_melted_beta_subset$threads_run <- paste0(data_melted_beta_subset$Threads, "_", data_melted_beta_subset$Run) #data_melted_beta_subset$Param <- NULL #data_melted_beta_subset$Threads <- NULL #data_melted_beta_subset$Run <- NULL #-----------------------STAR----------------------- data_melted_beta_subset_STAR <- data_melted_beta_subset %>% filter(Run == "STAR") data_melted_beta_subset_STAR_Param <- data_melted_beta_subset_STAR[1:4] data_melted_beta_subset_STAR_reads <- unique(data_melted_beta_subset_STAR[,c(1,6)]) data_melted_beta_subset_STAR_speed <- unique(data_melted_beta_subset_STAR[,c(1,4, 7)]) data_melted_beta_subset_STAR_Param$comb <- paste0(data_melted_beta_subset_STAR_Param$Param, "_", data_melted_beta_subset_STAR_Param$Threads, "_", "STAR") data_melted_beta_subset_STAR_Param$Param <- NULL data_melted_beta_subset_STAR_Param$Threads <- NULL data_melted_beta_subset_STAR_Param <- as.data.frame(data_melted_beta_subset_STAR_Param) data_melted_beta_subset_STAR_Param_spread <- spread(data_melted_beta_subset_STAR_Param, key = comb, value = Value) data_melted_beta_subset_STAR_Param_spread <- data_melted_beta_subset_STAR_Param_spread %>% dplyr::group_by(Sample) %>% dplyr::summarise_all(purrr::discard, is.na) data_melted_beta_subset_STAR_Param_spread <- as.data.frame(data_melted_beta_subset_STAR_Param_spread) data_melted_beta_subset_STAR_speed_spread <- spread(data_melted_beta_subset_STAR_speed, key = Threads, value = Mapping_speed_Million_of_reads_per_hour) colnames(data_melted_beta_subset_STAR_speed_spread) colnames(data_melted_beta_subset_STAR_speed_spread) <- c("Sample", "Mapping_Speed_8threads_STAR", "Mapping_Speed_16threads_STAR", "Mapping_Speed_32threads_STAR") STAR_tab_subset <- inner_join(data_melted_beta_subset_STAR_reads, data_melted_beta_subset_STAR_Param_spread, by = "Sample") STAR_tab_subset <- inner_join(STAR_tab_subset, data_melted_beta_subset_STAR_speed_spread, by = "Sample") STAR_tab_subset <- STAR_tab_subset[,c(1,2,3,4,5,9,10,11, 6,7,8)] #----------------------STAR_WASP----------------------- data_melted_beta_subset_STAR_WASP <- data_melted_beta_subset %>% filter(Run == "STAR_WASP") data_melted_beta_subset_STAR_WASP_Param <- data_melted_beta_subset_STAR_WASP[1:4] data_melted_beta_subset_STAR_WASP_reads <- unique(data_melted_beta_subset_STAR_WASP[,c(1,6)]) data_melted_beta_subset_STAR_WASP_speed <- unique(data_melted_beta_subset_STAR_WASP[,c(1,4, 7)]) data_melted_beta_subset_STAR_WASP_Param$comb <- paste0(data_melted_beta_subset_STAR_WASP_Param$Param, "_", data_melted_beta_subset_STAR_WASP_Param$Threads, "_", "STAR_WASP") data_melted_beta_subset_STAR_WASP_Param$Param <- NULL data_melted_beta_subset_STAR_WASP_Param$Threads <- NULL data_melted_beta_subset_STAR_WASP_Param <- as.data.frame(data_melted_beta_subset_STAR_WASP_Param) rownames(data_melted_beta_subset_STAR_WASP_Param) <- 1:nrow(data_melted_beta_subset_STAR_WASP_Param) data_melted_beta_subset_STAR_WASP_Param$id <- 1:nrow(data_melted_beta_subset_STAR_WASP_Param) data_melted_beta_subset_STAR_WASP_Param_spread <- spread(data_melted_beta_subset_STAR_WASP_Param, key = comb, value = Value) data_melted_beta_subset_STAR_WASP_Param_spread$id <- NULL data_melted_beta_subset_STAR_WASP_Param_spread <- data_melted_beta_subset_STAR_WASP_Param_spread %>% dplyr::group_by(Sample) %>% dplyr::summarise_all(purrr::discard, is.na) data_melted_beta_subset_STAR_WASP_Param_spread <- as.data.frame(data_melted_beta_subset_STAR_WASP_Param_spread) data_melted_beta_subset_STAR_WASP_speed_spread <- spread(data_melted_beta_subset_STAR_WASP_speed, key = Threads, value = Mapping_speed_Million_of_reads_per_hour) colnames(data_melted_beta_subset_STAR_WASP_speed_spread) colnames(data_melted_beta_subset_STAR_WASP_speed_spread) <- c("Sample", "Mapping_Speed_8threads_STAR_WASP", "Mapping_Speed_16threads_STAR_WASP", "Mapping_Speed_32threads_STAR_WASP") STAR_WASP_tab_subset <- inner_join(data_melted_beta_subset_STAR_WASP_Param_spread, data_melted_beta_subset_STAR_WASP_speed_spread, by = "Sample") STAR_WASP_tab_subset <- STAR_WASP_tab_subset[,c(1,2,3,4,8,9,10, 5,6,7)] #-------------------WASP------------------------------- data_melted_beta_subset_WASP <- data_melted_beta_subset %>% filter(Run == "WASP") data_melted_beta_subset_WASP_Param <- data_melted_beta_subset_WASP[1:4] data_melted_beta_subset_WASP_reads <- unique(data_melted_beta_subset_WASP[,c(1,6)]) data_melted_beta_subset_WASP_speed <- unique(data_melted_beta_subset_WASP[,c(1,4, 7)]) data_melted_beta_subset_WASP_Param$comb <- paste0(data_melted_beta_subset_WASP_Param$Param, "_", data_melted_beta_subset_WASP_Param$Threads, "_", "WASP") data_melted_beta_subset_WASP_Param$Param <- NULL data_melted_beta_subset_WASP_Param$Threads <- NULL data_melted_beta_subset_WASP_Param <- as.data.frame(data_melted_beta_subset_WASP_Param) rownames(data_melted_beta_subset_WASP_Param) <- 1:nrow(data_melted_beta_subset_WASP_Param) data_melted_beta_subset_WASP_Param$id <- 1:nrow(data_melted_beta_subset_WASP_Param) data_melted_beta_subset_WASP_Param_spread <- spread(data_melted_beta_subset_WASP_Param, key = comb, value = Value) data_melted_beta_subset_WASP_Param_spread$id <- NULL data_melted_beta_subset_WASP_Param_spread <- data_melted_beta_subset_WASP_Param_spread %>% dplyr::group_by(Sample) %>% dplyr::summarise_all(purrr::discard, is.na) data_melted_beta_subset_WASP_Param_spread <- as.data.frame(data_melted_beta_subset_WASP_Param_spread) data_melted_beta_subset_WASP_speed_spread <- spread(data_melted_beta_subset_WASP_speed, key = Threads, value = Mapping_speed_Million_of_reads_per_hour) colnames(data_melted_beta_subset_WASP_speed_spread) colnames(data_melted_beta_subset_WASP_speed_spread) <- c("Sample", "Mapping_Speed_8threads_WASP", "Mapping_Speed_16threads_WASP", "Mapping_Speed_32threads_WASP") WASP_tab_subset <- inner_join(data_melted_beta_subset_WASP_Param_spread, data_melted_beta_subset_WASP_speed_spread, by = "Sample") WASP_tab_subset <- WASP_tab_subset[,c(1,2,3,4,8,9,10, 5,6,7)] STAR_tab_subset <- as.data.frame(STAR_tab_subset) STAR_WASP_tab_subset <- as.data.frame(STAR_WASP_tab_subset) WASP_tab_subset <- as.data.frame(WASP_tab_subset) n_way_merger0 <- merge(STAR_tab_subset, STAR_WASP_tab_subset, by = "Sample") n_way_merger <- merge(n_way_merger0, WASP_tab_subset, by = "Sample") n_way_merger$Size_holder1 <- "" n_way_merger$Size_holder2 <- "" colnames(n_way_merger) #[1] "Sample" "Number_of_input_reads" "Memory_16threads_STAR" "Memory_32threads_STAR" "Memory_8threads_STAR" #[6] "Mapping_Speed_8threads_STAR" "Mapping_Speed_16threads_STAR" "Mapping_Speed_32threads_STAR" "Wall_Clock_16threads_STAR" "Wall_Clock_32threads_STAR" #[11] "Wall_Clock_8threads_STAR" "Memory_16threads_STAR_WASP" "Memory_32threads_STAR_WASP" "Memory_8threads_STAR_WASP" "Mapping_Speed_8threads_STAR_WASP" #[16] "Mapping_Speed_16threads_STAR_WASP" "Mapping_Speed_32threads_STAR_WASP" "Wall_Clock_16threads_STAR_WASP" "Wall_Clock_32threads_STAR_WASP" "Wall_Clock_8threads_STAR_WASP" #[21] "Memory_16threads_WASP" "Memory_32threads_WASP" "Memory_8threads_WASP" "Mapping_Speed_8threads_WASP" "Mapping_Speed_16threads_WASP" #[26] "Mapping_Speed_32threads_WASP" "Wall_Clock_16threads_WASP" "Wall_Clock_32threads_WASP" "Wall_Clock_8threads_WASP" "Size_holder1" #[31] "Size_holder2" #n_way_merger <- n_way_merger[] #reorg cols based on table - nreads | Ram |speed | Clock knitr::kable(n_way_merger, "latex")